| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 155, | |
| "global_step": 620, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.7582, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "eval_loss": 2.128222703933716, | |
| "eval_runtime": 1.7527, | |
| "eval_samples_per_second": 19.398, | |
| "eval_steps_per_second": 19.398, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.0767, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.875, | |
| "learning_rate": 6e-06, | |
| "loss": 2.074, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.21875, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.9594, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.75, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1593, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.5345, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 1.7973, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.875, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.0634, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.5362, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 7.125, | |
| "learning_rate": 2e-05, | |
| "loss": 2.2476, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 1.8936, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.5785, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 1.5667, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 1.9219, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 3e-05, | |
| "loss": 2.0164, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.0587, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 1.9962, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 28.25, | |
| "learning_rate": 3.6e-05, | |
| "loss": 2.2685, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 7.5, | |
| "learning_rate": 3.8e-05, | |
| "loss": 2.3435, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 4e-05, | |
| "loss": 2.3058, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 4.2e-05, | |
| "loss": 2.1123, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 2.8498, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 6.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.9676, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 7.625, | |
| "learning_rate": 4.8e-05, | |
| "loss": 2.3663, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 5e-05, | |
| "loss": 1.8029, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 17.125, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 2.0883, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 1.8337, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 2.0822, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 5.8e-05, | |
| "loss": 2.3772, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 5.875, | |
| "learning_rate": 6e-05, | |
| "loss": 1.8399, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 6.2e-05, | |
| "loss": 1.7067, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.8348, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.15625, | |
| "learning_rate": 6.6e-05, | |
| "loss": 1.9547, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 2.3197, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 7.5, | |
| "learning_rate": 7e-05, | |
| "loss": 2.2976, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 7.2e-05, | |
| "loss": 2.2088, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 7.4e-05, | |
| "loss": 1.8717, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 7.6e-05, | |
| "loss": 2.8409, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 7.875, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 2.1854, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 5.375, | |
| "learning_rate": 8e-05, | |
| "loss": 1.741, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 8.2e-05, | |
| "loss": 2.3502, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 8.4e-05, | |
| "loss": 1.6967, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 8.6e-05, | |
| "loss": 1.9294, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.834, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 11.125, | |
| "learning_rate": 9e-05, | |
| "loss": 1.4147, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 1.6542, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 9.0, | |
| "learning_rate": 9.4e-05, | |
| "loss": 2.4612, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.8718, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 9.8e-05, | |
| "loss": 2.3953, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 0.0001, | |
| "loss": 2.0245, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 8.875, | |
| "learning_rate": 0.00010200000000000001, | |
| "loss": 2.251, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 0.00010400000000000001, | |
| "loss": 2.2304, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 8.625, | |
| "learning_rate": 0.00010600000000000002, | |
| "loss": 2.0407, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 8.25, | |
| "learning_rate": 0.00010800000000000001, | |
| "loss": 2.4401, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 1.9831, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 0.00011200000000000001, | |
| "loss": 1.9876, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 0.00011399999999999999, | |
| "loss": 2.0943, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 0.000116, | |
| "loss": 2.0148, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 0.000118, | |
| "loss": 1.8346, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 0.00012, | |
| "loss": 2.0344, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 0.000122, | |
| "loss": 1.8906, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 0.000124, | |
| "loss": 2.4039, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 10.3125, | |
| "learning_rate": 0.000126, | |
| "loss": 2.052, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 13.3125, | |
| "learning_rate": 0.00012800000000000002, | |
| "loss": 2.7162, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.75, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 1.6805, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 13.375, | |
| "learning_rate": 0.000132, | |
| "loss": 2.9215, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 0.000134, | |
| "loss": 2.0593, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 0.00013600000000000003, | |
| "loss": 2.1428, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 7.0, | |
| "learning_rate": 0.000138, | |
| "loss": 2.4703, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 10.75, | |
| "learning_rate": 0.00014, | |
| "loss": 2.5697, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 9.625, | |
| "learning_rate": 0.000142, | |
| "loss": 2.8569, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 0.000144, | |
| "loss": 2.3486, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 0.000146, | |
| "loss": 2.2554, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 10.0, | |
| "learning_rate": 0.000148, | |
| "loss": 2.2662, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 2.4359, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 0.000152, | |
| "loss": 3.3211, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 13.375, | |
| "learning_rate": 0.000154, | |
| "loss": 2.9317, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 0.00015600000000000002, | |
| "loss": 2.7471, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 6.5, | |
| "learning_rate": 0.00015800000000000002, | |
| "loss": 2.2128, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 15.25, | |
| "learning_rate": 0.00016, | |
| "loss": 2.613, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 0.000162, | |
| "loss": 3.2154, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 7.5, | |
| "learning_rate": 0.000164, | |
| "loss": 2.3345, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 13.25, | |
| "learning_rate": 0.000166, | |
| "loss": 2.1628, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 11.0, | |
| "learning_rate": 0.000168, | |
| "loss": 2.4309, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 9.25, | |
| "learning_rate": 0.00017, | |
| "loss": 2.8301, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 0.000172, | |
| "loss": 2.3429, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 20.0, | |
| "learning_rate": 0.000174, | |
| "loss": 3.0507, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 9.75, | |
| "learning_rate": 0.00017600000000000002, | |
| "loss": 2.5522, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 0.00017800000000000002, | |
| "loss": 2.22, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 11.5, | |
| "learning_rate": 0.00018, | |
| "loss": 2.4792, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 19.375, | |
| "learning_rate": 0.000182, | |
| "loss": 2.8168, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 10.0, | |
| "learning_rate": 0.00018400000000000003, | |
| "loss": 2.7499, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 11.125, | |
| "learning_rate": 0.00018600000000000002, | |
| "loss": 2.12, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 0.000188, | |
| "loss": 2.5596, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 0.00019, | |
| "loss": 2.2917, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.000192, | |
| "loss": 2.5138, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 0.000194, | |
| "loss": 2.6197, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 0.000196, | |
| "loss": 2.4397, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 9.375, | |
| "learning_rate": 0.00019800000000000002, | |
| "loss": 2.3659, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 46.5, | |
| "learning_rate": 0.0002, | |
| "loss": 2.0891, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 0.00019999817500473724, | |
| "loss": 2.3711, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 0.00019999270008556108, | |
| "loss": 2.5669, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 9.375, | |
| "learning_rate": 0.00019998357544230558, | |
| "loss": 2.7071, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 9.0, | |
| "learning_rate": 0.00019997080140801932, | |
| "loss": 2.0942, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 0.00019995437844895334, | |
| "loss": 2.6619, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 9.375, | |
| "learning_rate": 0.00019993430716454413, | |
| "loss": 2.0561, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 0.00019991058828739165, | |
| "loss": 2.5201, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 17.375, | |
| "learning_rate": 0.00019988322268323268, | |
| "loss": 2.6833, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 0.00019985221135090914, | |
| "loss": 2.3537, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 0.00019981755542233177, | |
| "loss": 2.2725, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 44.25, | |
| "learning_rate": 0.00019977925616243862, | |
| "loss": 2.5759, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 18.125, | |
| "learning_rate": 0.00019973731496914914, | |
| "loss": 2.7988, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 10.0, | |
| "learning_rate": 0.0001996917333733128, | |
| "loss": 2.4719, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 0.00019964251303865362, | |
| "loss": 3.1389, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 16.25, | |
| "learning_rate": 0.00019958965576170908, | |
| "loss": 2.5832, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 11.125, | |
| "learning_rate": 0.00019953316347176488, | |
| "loss": 2.8191, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 0.00019947303823078416, | |
| "loss": 2.6941, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 11.875, | |
| "learning_rate": 0.00019940928223333252, | |
| "loss": 2.3098, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 11.375, | |
| "learning_rate": 0.0001993418978064979, | |
| "loss": 2.6137, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 16.375, | |
| "learning_rate": 0.0001992708874098054, | |
| "loss": 4.0113, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 7.25, | |
| "learning_rate": 0.00019919625363512786, | |
| "loss": 2.4582, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 19.5, | |
| "learning_rate": 0.00019911799920659093, | |
| "loss": 2.1897, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 20.25, | |
| "learning_rate": 0.00019903612698047383, | |
| "loss": 2.9714, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 18.375, | |
| "learning_rate": 0.0001989506399451051, | |
| "loss": 3.2223, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 189.0, | |
| "learning_rate": 0.00019886154122075343, | |
| "loss": 8.0224, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 19.75, | |
| "learning_rate": 0.00019876883405951377, | |
| "loss": 2.276, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 0.00019867252184518878, | |
| "loss": 2.263, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 129.0, | |
| "learning_rate": 0.0001985726080931651, | |
| "loss": 9.2328, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 134.0, | |
| "learning_rate": 0.00019846909645028523, | |
| "loss": 7.3449, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 25.625, | |
| "learning_rate": 0.00019836199069471437, | |
| "loss": 2.7, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 12.625, | |
| "learning_rate": 0.0001982512947358024, | |
| "loss": 2.18, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 27.0, | |
| "learning_rate": 0.00019813701261394136, | |
| "loss": 2.9489, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 13.4375, | |
| "learning_rate": 0.00019801914850041784, | |
| "loss": 2.2871, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 16.625, | |
| "learning_rate": 0.00019789770669726087, | |
| "loss": 2.6805, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 136.0, | |
| "learning_rate": 0.00019777269163708468, | |
| "loss": 2.0789, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 23.25, | |
| "learning_rate": 0.00019764410788292722, | |
| "loss": 2.7603, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 53.75, | |
| "learning_rate": 0.00019751196012808325, | |
| "loss": 2.6992, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 84.5, | |
| "learning_rate": 0.00019737625319593335, | |
| "loss": 8.1541, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 320.0, | |
| "learning_rate": 0.00019723699203976766, | |
| "loss": 8.7045, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 1192.0, | |
| "learning_rate": 0.0001970941817426052, | |
| "loss": 3.8857, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 202.0, | |
| "learning_rate": 0.00019694782751700828, | |
| "loss": 3.4135, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 14.1875, | |
| "learning_rate": 0.00019679793470489228, | |
| "loss": 2.4865, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 18.625, | |
| "learning_rate": 0.00019664450877733062, | |
| "loss": 2.2073, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 22.625, | |
| "learning_rate": 0.00019648755533435518, | |
| "loss": 2.4245, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 92.0, | |
| "learning_rate": 0.00019632708010475165, | |
| "loss": 2.2213, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 25.875, | |
| "learning_rate": 0.00019616308894585078, | |
| "loss": 2.3082, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 4512.0, | |
| "learning_rate": 0.0001959955878433143, | |
| "loss": 7.3644, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 0.00019582458291091663, | |
| "loss": 2.4085, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 50.25, | |
| "learning_rate": 0.00019565008039032158, | |
| "loss": 9.1058, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 0.00019547208665085457, | |
| "loss": 2.3966, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 15.4375, | |
| "learning_rate": 0.0001952906081892703, | |
| "loss": 2.4106, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1448.0, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 3.339, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 161.0, | |
| "learning_rate": 0.0001949172237224867, | |
| "loss": 3.5134, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 486.0, | |
| "learning_rate": 0.00019472533134578507, | |
| "loss": 7.866, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 17.75, | |
| "learning_rate": 0.00019452998150346401, | |
| "loss": 2.6905, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 4.079593658447266, | |
| "eval_runtime": 1.7475, | |
| "eval_samples_per_second": 19.457, | |
| "eval_steps_per_second": 19.457, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 296.0, | |
| "learning_rate": 0.0001943311813257743, | |
| "loss": 2.2737, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 93.5, | |
| "learning_rate": 0.00019412893806890357, | |
| "loss": 2.3946, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 36.0, | |
| "learning_rate": 0.00019392325911471155, | |
| "loss": 2.5526, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 114.0, | |
| "learning_rate": 0.00019371415197046052, | |
| "loss": 2.3686, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1016.0, | |
| "learning_rate": 0.0001935016242685415, | |
| "loss": 4.7788, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 86.5, | |
| "learning_rate": 0.00019328568376619543, | |
| "loss": 2.9732, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 205.0, | |
| "learning_rate": 0.00019306633834523024, | |
| "loss": 2.886, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 41.0, | |
| "learning_rate": 0.00019284359601173294, | |
| "loss": 2.3179, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 46.0, | |
| "learning_rate": 0.00019261746489577765, | |
| "loss": 2.6493, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 16.625, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 3.2452, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 17.375, | |
| "learning_rate": 0.0001921550694549393, | |
| "loss": 2.8346, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 16.5, | |
| "learning_rate": 0.000191918822007446, | |
| "loss": 2.4778, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 262.0, | |
| "learning_rate": 0.00019167921953165825, | |
| "loss": 3.5156, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 59.5, | |
| "learning_rate": 0.0001914362707730437, | |
| "loss": 2.7138, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 25.625, | |
| "learning_rate": 0.00019118998459920902, | |
| "loss": 2.9188, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 432.0, | |
| "learning_rate": 0.00019094036999957624, | |
| "loss": 4.9694, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 43.0, | |
| "learning_rate": 0.00019068743608505455, | |
| "loss": 2.2251, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 36.0, | |
| "learning_rate": 0.00019043119208770793, | |
| "loss": 2.3021, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 11.5, | |
| "learning_rate": 0.00019017164736041795, | |
| "loss": 2.6769, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 21.875, | |
| "learning_rate": 0.00018990881137654258, | |
| "loss": 2.9547, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 12.5, | |
| "learning_rate": 0.00018964269372957038, | |
| "loss": 2.5707, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 235.0, | |
| "learning_rate": 0.0001893733041327702, | |
| "loss": 2.3129, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 10.875, | |
| "learning_rate": 0.0001891006524188368, | |
| "loss": 2.3162, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 32.75, | |
| "learning_rate": 0.0001888247485395319, | |
| "loss": 2.3063, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 17.75, | |
| "learning_rate": 0.000188545602565321, | |
| "loss": 3.1659, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 0.00018826322468500566, | |
| "loss": 2.4882, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 35.25, | |
| "learning_rate": 0.00018797762520535177, | |
| "loss": 2.8105, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 9.0, | |
| "learning_rate": 0.00018768881455071332, | |
| "loss": 2.2759, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 70.0, | |
| "learning_rate": 0.0001873968032626518, | |
| "loss": 2.8899, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 7.28125, | |
| "learning_rate": 0.00018710160199955156, | |
| "loss": 2.6619, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 9.125, | |
| "learning_rate": 0.00018680322153623075, | |
| "loss": 2.6123, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 17.375, | |
| "learning_rate": 0.000186501672763548, | |
| "loss": 2.5625, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 0.00018619696668800492, | |
| "loss": 2.6897, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 0.00018588911443134448, | |
| "loss": 3.0509, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 27.75, | |
| "learning_rate": 0.00018557812723014476, | |
| "loss": 2.5462, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 9.5, | |
| "learning_rate": 0.00018526401643540922, | |
| "loss": 2.0449, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 15.9375, | |
| "learning_rate": 0.0001849467935121521, | |
| "loss": 3.0498, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 0.00018462647003898006, | |
| "loss": 3.1474, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 0.00018430305770766948, | |
| "loss": 2.4962, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 2464.0, | |
| "learning_rate": 0.0001839765683227398, | |
| "loss": 12.6715, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 13.0, | |
| "learning_rate": 0.00018364701380102266, | |
| "loss": 2.5459, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 43.25, | |
| "learning_rate": 0.00018331440617122696, | |
| "loss": 2.8457, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 0.00018297875757349952, | |
| "loss": 2.9238, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 19.5, | |
| "learning_rate": 0.00018264008025898248, | |
| "loss": 2.4803, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 16.0, | |
| "learning_rate": 0.00018229838658936564, | |
| "loss": 2.7507, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 17.0, | |
| "learning_rate": 0.00018195368903643563, | |
| "loss": 2.4951, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 0.0001816060001816205, | |
| "loss": 1.9929, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 14.25, | |
| "learning_rate": 0.00018125533271553043, | |
| "loss": 2.7251, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 45.5, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 2.2007, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 8.5, | |
| "learning_rate": 0.0001805451132550946, | |
| "loss": 2.6761, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 0.00018018558718369186, | |
| "loss": 2.1643, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 13.75, | |
| "learning_rate": 0.00017982313434595406, | |
| "loss": 1.9931, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 7.875, | |
| "learning_rate": 0.00017945776797137543, | |
| "loss": 2.9199, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 24.75, | |
| "learning_rate": 0.00017908950139579406, | |
| "loss": 3.1686, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 6240.0, | |
| "learning_rate": 0.00017871834806090501, | |
| "loss": 4.8329, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 0.0001783443215137699, | |
| "loss": 2.3425, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 11.125, | |
| "learning_rate": 0.00017796743540632223, | |
| "loss": 7.5991, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 0.00017758770349486923, | |
| "loss": 2.5081, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 12.25, | |
| "learning_rate": 0.00017720513963958968, | |
| "loss": 2.3669, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 31.875, | |
| "learning_rate": 0.00017681975780402807, | |
| "loss": 3.2162, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 13.8125, | |
| "learning_rate": 0.00017643157205458483, | |
| "loss": 2.8832, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 0.0001760405965600031, | |
| "loss": 2.1019, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 0.00017564684559085136, | |
| "loss": 2.1741, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 88.5, | |
| "learning_rate": 0.00017525033351900268, | |
| "loss": 2.875, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 15.5, | |
| "learning_rate": 0.00017485107481711012, | |
| "loss": 2.9173, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 18.875, | |
| "learning_rate": 0.00017444908405807845, | |
| "loss": 2.77, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 114.5, | |
| "learning_rate": 0.00017404437591453235, | |
| "loss": 2.5471, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 19.5, | |
| "learning_rate": 0.00017363696515828062, | |
| "loss": 2.4391, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 14.0625, | |
| "learning_rate": 0.00017322686665977737, | |
| "loss": 2.4229, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 33.5, | |
| "learning_rate": 0.00017281409538757883, | |
| "loss": 2.9198, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 25.125, | |
| "learning_rate": 0.00017239866640779745, | |
| "loss": 2.8782, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 50.25, | |
| "learning_rate": 0.0001719805948835515, | |
| "loss": 3.3913, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 0.00017155989607441213, | |
| "loss": 2.4202, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 398.0, | |
| "learning_rate": 0.00017113658533584594, | |
| "loss": 2.6242, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 22.0, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 2.3384, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 9.875, | |
| "learning_rate": 0.00017028218996841172, | |
| "loss": 2.4174, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 0.00016985113652489374, | |
| "loss": 2.3576, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 20.0, | |
| "learning_rate": 0.00016941753352151055, | |
| "loss": 2.7567, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 19.75, | |
| "learning_rate": 0.00016898139678473076, | |
| "loss": 2.6665, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 8.125, | |
| "learning_rate": 0.00016854274223350397, | |
| "loss": 2.5973, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 0.00016810158587867973, | |
| "loss": 2.7815, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 10.25, | |
| "learning_rate": 0.00016765794382242314, | |
| "loss": 2.6786, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 5.75, | |
| "learning_rate": 0.00016721183225762727, | |
| "loss": 1.876, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 0.00016676326746732195, | |
| "loss": 1.9815, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 0.00016631226582407952, | |
| "loss": 2.2768, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 9.0, | |
| "learning_rate": 0.00016585884378941725, | |
| "loss": 2.6424, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 0.00016540301791319645, | |
| "loss": 2.2763, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 9.0, | |
| "learning_rate": 0.00016494480483301836, | |
| "loss": 2.6912, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 9.25, | |
| "learning_rate": 0.00016448422127361706, | |
| "loss": 2.0026, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 13.875, | |
| "learning_rate": 0.00016402128404624882, | |
| "loss": 2.4934, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 7.53125, | |
| "learning_rate": 0.00016355601004807856, | |
| "loss": 2.2699, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 9.25, | |
| "learning_rate": 0.00016308841626156307, | |
| "loss": 2.2252, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 0.00016261851975383137, | |
| "loss": 2.9339, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 8.625, | |
| "learning_rate": 0.00016214633767606143, | |
| "loss": 3.2039, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 0.00016167188726285434, | |
| "loss": 2.7348, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 0.0001611951858316052, | |
| "loss": 2.8089, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 22.75, | |
| "learning_rate": 0.00016071625078187114, | |
| "loss": 3.457, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 7.875, | |
| "learning_rate": 0.00016023509959473605, | |
| "loss": 2.8804, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 11.25, | |
| "learning_rate": 0.00015975174983217275, | |
| "loss": 2.3933, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 0.0001592662191364017, | |
| "loss": 2.4153, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 2.9298, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 9.25, | |
| "learning_rate": 0.00015828868591149104, | |
| "loss": 2.7037, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 0.0001577967190622215, | |
| "loss": 2.8552, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 83.0, | |
| "learning_rate": 0.00015730264263818212, | |
| "loss": 3.3196, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 0.00015680647467311557, | |
| "loss": 2.6933, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 9.875, | |
| "learning_rate": 0.00015630823327710558, | |
| "loss": 2.8295, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 400.0, | |
| "learning_rate": 0.00015580793663591585, | |
| "loss": 7.8334, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 0.0001553056030103264, | |
| "loss": 3.0095, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 14.25, | |
| "learning_rate": 0.00015480125073546704, | |
| "loss": 2.8319, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 0.0001542948982201479, | |
| "loss": 2.7952, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 63.5, | |
| "learning_rate": 0.00015378656394618787, | |
| "loss": 3.3161, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 0.00015327626646773976, | |
| "loss": 2.3066, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 13.4375, | |
| "learning_rate": 0.0001527640244106133, | |
| "loss": 2.9634, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 0.0001522498564715949, | |
| "loss": 2.9318, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 0.00015173378141776568, | |
| "loss": 2.6438, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 520.0, | |
| "learning_rate": 0.00015121581808581622, | |
| "loss": 10.9186, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.00015069598538135906, | |
| "loss": 2.6323, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 8.125, | |
| "learning_rate": 0.00015017430227823864, | |
| "loss": 2.9288, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 14336.0, | |
| "learning_rate": 0.0001496507878178388, | |
| "loss": 16.6016, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 11.0, | |
| "learning_rate": 0.00014912546110838775, | |
| "loss": 3.2856, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 13.0, | |
| "learning_rate": 0.0001485983413242606, | |
| "loss": 2.1977, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 70.5, | |
| "learning_rate": 0.00014806944770527958, | |
| "loss": 13.3843, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 0.00014753879955601163, | |
| "loss": 3.659, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 7.3125, | |
| "learning_rate": 0.00014700641624506392, | |
| "loss": 2.1687, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 10.75, | |
| "learning_rate": 0.00014647231720437686, | |
| "loss": 2.7034, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 100.0, | |
| "learning_rate": 0.00014593652192851486, | |
| "loss": 8.6095, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.00014539904997395468, | |
| "loss": 2.0712, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 414.0, | |
| "learning_rate": 0.00014485992095837177, | |
| "loss": 1.984, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 0.00014431915455992414, | |
| "loss": 2.2412, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 0.00014377677051653404, | |
| "loss": 2.4154, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 0.00014323278862516775, | |
| "loss": 2.6303, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.375, | |
| "learning_rate": 0.00014268722874111265, | |
| "loss": 2.42, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 0.00014214011077725292, | |
| "loss": 2.403, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 10.125, | |
| "learning_rate": 0.00014159145470334235, | |
| "loss": 2.8993, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 0.0001410412805452757, | |
| "loss": 2.2136, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 0.00014048960838435753, | |
| "loss": 1.9811, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 7.625, | |
| "learning_rate": 0.00013993645835656953, | |
| "loss": 2.8393, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 6.875, | |
| "learning_rate": 0.00013938185065183532, | |
| "loss": 2.2938, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 10.75, | |
| "learning_rate": 0.0001388258055132835, | |
| "loss": 2.9148, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 0.000138268343236509, | |
| "loss": 2.8022, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 10.125, | |
| "learning_rate": 0.00013770948416883205, | |
| "loss": 3.0303, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 0.00013714924870855571, | |
| "loss": 2.5784, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 0.00013658765730422125, | |
| "loss": 2.2308, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 13.6875, | |
| "learning_rate": 0.00013602473045386165, | |
| "loss": 2.5253, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 14.9375, | |
| "learning_rate": 0.00013546048870425356, | |
| "loss": 2.809, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 11.375, | |
| "learning_rate": 0.0001348949526501675, | |
| "loss": 3.2099, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 14.75, | |
| "learning_rate": 0.00013432814293361584, | |
| "loss": 2.0988, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 106.5, | |
| "learning_rate": 0.00013376008024309948, | |
| "loss": 2.4118, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 0.00013319078531285285, | |
| "loss": 3.1705, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 7.5, | |
| "learning_rate": 0.00013262027892208694, | |
| "loss": 2.5198, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 0.00013204858189423097, | |
| "loss": 2.875, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 0.00013147571509617228, | |
| "loss": 2.9605, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 2.2224, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 0.00013032655586971552, | |
| "loss": 2.577, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 14.25, | |
| "learning_rate": 0.00012975030538552032, | |
| "loss": 2.9887, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 2.833022356033325, | |
| "eval_runtime": 1.7613, | |
| "eval_samples_per_second": 19.304, | |
| "eval_steps_per_second": 19.304, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 7.90625, | |
| "learning_rate": 0.0001291729690179972, | |
| "loss": 2.6105, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 25.125, | |
| "learning_rate": 0.00012859456783986893, | |
| "loss": 2.7618, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 8.0, | |
| "learning_rate": 0.00012801512296272368, | |
| "loss": 3.0351, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 6.1875, | |
| "learning_rate": 0.0001274346555362446, | |
| "loss": 2.5821, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.0001268531867474377, | |
| "loss": 2.5642, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 24.5, | |
| "learning_rate": 0.0001262707378198587, | |
| "loss": 2.1907, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 5.75, | |
| "learning_rate": 0.00012568733001283827, | |
| "loss": 2.3074, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 11.25, | |
| "learning_rate": 0.00012510298462070619, | |
| "loss": 2.5847, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.00012451772297201376, | |
| "loss": 2.3077, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 0.0001239315664287558, | |
| "loss": 2.4337, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 0.00012334453638559057, | |
| "loss": 2.3457, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 6.34375, | |
| "learning_rate": 0.000122756654269059, | |
| "loss": 2.7569, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 0.00012216794153680274, | |
| "loss": 2.8255, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 6.875, | |
| "learning_rate": 0.00012157841967678063, | |
| "loss": 2.4809, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 8.375, | |
| "learning_rate": 0.00012098811020648475, | |
| "loss": 2.7188, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 0.00012039703467215488, | |
| "loss": 2.7395, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 6.9375, | |
| "learning_rate": 0.00011980521464799198, | |
| "loss": 2.5989, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 0.00011921267173537086, | |
| "loss": 2.5155, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 0.00011861942756205169, | |
| "loss": 2.4556, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 6.96875, | |
| "learning_rate": 0.0001180255037813906, | |
| "loss": 2.2609, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 0.00011743092207154929, | |
| "loss": 2.6818, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 0.00011683570413470383, | |
| "loss": 2.5213, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 0.00011623987169625261, | |
| "loss": 2.513, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 0.0001156434465040231, | |
| "loss": 2.8551, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 0.00011504645032747832, | |
| "loss": 2.336, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 6.5, | |
| "learning_rate": 0.00011444890495692213, | |
| "loss": 2.4159, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 7.59375, | |
| "learning_rate": 0.00011385083220270401, | |
| "loss": 2.8159, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 0.00011325225389442277, | |
| "loss": 3.1515, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 0.00011265319188012994, | |
| "loss": 2.5154, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 0.0001120536680255323, | |
| "loss": 2.9747, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 5.75, | |
| "learning_rate": 0.00011145370421319377, | |
| "loss": 2.2161, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 5.0, | |
| "learning_rate": 0.00011085332234173664, | |
| "loss": 2.8727, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 6.5, | |
| "learning_rate": 0.00011025254432504233, | |
| "loss": 2.5946, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 0.00010965139209145152, | |
| "loss": 2.898, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 0.0001090498875829638, | |
| "loss": 2.4455, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 6.25, | |
| "learning_rate": 0.00010844805275443673, | |
| "loss": 2.4731, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 0.0001078459095727845, | |
| "loss": 2.1894, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 13.125, | |
| "learning_rate": 0.00010724348001617625, | |
| "loss": 2.7624, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 35.5, | |
| "learning_rate": 0.00010664078607323367, | |
| "loss": 2.284, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 0.00010603784974222861, | |
| "loss": 2.2169, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 10.875, | |
| "learning_rate": 0.00010543469303028002, | |
| "loss": 2.6135, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 5.78125, | |
| "learning_rate": 0.00010483133795255071, | |
| "loss": 2.6733, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 5.875, | |
| "learning_rate": 0.0001042278065314439, | |
| "loss": 2.2963, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 5.625, | |
| "learning_rate": 0.00010362412079579924, | |
| "loss": 2.2819, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 6.0, | |
| "learning_rate": 0.0001030203027800889, | |
| "loss": 2.7232, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 17.875, | |
| "learning_rate": 0.00010241637452361323, | |
| "loss": 2.8421, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 24.875, | |
| "learning_rate": 0.0001018123580696964, | |
| "loss": 2.554, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 5.5, | |
| "learning_rate": 0.00010120827546488174, | |
| "loss": 2.4021, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 0.00010060414875812709, | |
| "loss": 2.0252, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 11.125, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2328, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 6.875, | |
| "learning_rate": 9.939585124187292e-05, | |
| "loss": 2.6733, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 9.879172453511827e-05, | |
| "loss": 2.2601, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 9.818764193030363e-05, | |
| "loss": 2.1966, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 33.25, | |
| "learning_rate": 9.75836254763868e-05, | |
| "loss": 2.4876, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 15.75, | |
| "learning_rate": 9.697969721991114e-05, | |
| "loss": 2.2825, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 9.63758792042008e-05, | |
| "loss": 2.8742, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 7.0, | |
| "learning_rate": 9.577219346855613e-05, | |
| "loss": 2.3392, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 9.516866204744931e-05, | |
| "loss": 2.4111, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 9.456530696971999e-05, | |
| "loss": 2.2599, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 9.396215025777139e-05, | |
| "loss": 2.8202, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 9.335921392676631e-05, | |
| "loss": 2.3614, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 9.275651998382377e-05, | |
| "loss": 1.7052, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 9.215409042721552e-05, | |
| "loss": 1.8829, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 9.155194724556331e-05, | |
| "loss": 2.2968, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 9.095011241703623e-05, | |
| "loss": 2.1253, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 5.125, | |
| "learning_rate": 9.034860790854849e-05, | |
| "loss": 2.3356, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 8.974745567495768e-05, | |
| "loss": 2.1902, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 8.914667765826338e-05, | |
| "loss": 2.7202, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 8.854629578680624e-05, | |
| "loss": 2.4657, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 8.79463319744677e-05, | |
| "loss": 2.5267, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 8.73468081198701e-05, | |
| "loss": 2.4513, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 8.674774610557728e-05, | |
| "loss": 2.2212, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 6.0, | |
| "learning_rate": 8.614916779729603e-05, | |
| "loss": 2.2433, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 8.55510950430779e-05, | |
| "loss": 2.5887, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 8.495354967252169e-05, | |
| "loss": 1.5424, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 8.435655349597689e-05, | |
| "loss": 2.4985, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 7.0, | |
| "learning_rate": 8.37601283037474e-05, | |
| "loss": 3.0591, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 5.25, | |
| "learning_rate": 8.316429586529615e-05, | |
| "loss": 2.5296, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 8.256907792845072e-05, | |
| "loss": 2.4503, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 8.197449621860943e-05, | |
| "loss": 1.951, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 7.0, | |
| "learning_rate": 8.138057243794833e-05, | |
| "loss": 2.2415, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 8.078732826462915e-05, | |
| "loss": 3.0526, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 8.019478535200806e-05, | |
| "loss": 2.0695, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 7.960296532784515e-05, | |
| "loss": 1.9639, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 7.901188979351526e-05, | |
| "loss": 2.5458, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.875, | |
| "learning_rate": 7.84215803232194e-05, | |
| "loss": 2.6561, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 7.78320584631973e-05, | |
| "loss": 1.9434, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.375, | |
| "learning_rate": 7.7243345730941e-05, | |
| "loss": 2.291, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 7.66554636144095e-05, | |
| "loss": 1.9882, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 6.40625, | |
| "learning_rate": 7.606843357124426e-05, | |
| "loss": 2.5872, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 7.548227702798624e-05, | |
| "loss": 2.3354, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 7.489701537929384e-05, | |
| "loss": 2.9243, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.875, | |
| "learning_rate": 7.431266998716171e-05, | |
| "loss": 2.5853, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 7.372926218014131e-05, | |
| "loss": 2.3285, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 7.314681325256232e-05, | |
| "loss": 2.222, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 7.256534446375542e-05, | |
| "loss": 2.165, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 7.198487703727632e-05, | |
| "loss": 2.1, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 6.75, | |
| "learning_rate": 7.14054321601311e-05, | |
| "loss": 2.7438, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 7.082703098200282e-05, | |
| "loss": 1.7364, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 7.024969461447972e-05, | |
| "loss": 3.0458, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 6.967344413028452e-05, | |
| "loss": 2.2354, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 2.6881, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 141.0, | |
| "learning_rate": 6.852428490382773e-05, | |
| "loss": 10.2024, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 35.5, | |
| "learning_rate": 6.795141810576906e-05, | |
| "loss": 2.0233, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 5.625, | |
| "learning_rate": 6.73797210779131e-05, | |
| "loss": 2.0324, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 6.680921468714719e-05, | |
| "loss": 1.9068, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 6.623991975690051e-05, | |
| "loss": 2.4164, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 6.567185706638417e-05, | |
| "loss": 2.4648, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 6.510504734983249e-05, | |
| "loss": 1.9487, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 6.453951129574644e-05, | |
| "loss": 1.5947, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 6.397526954613839e-05, | |
| "loss": 2.0581, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 6.341234269577879e-05, | |
| "loss": 2.4201, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 38.25, | |
| "learning_rate": 6.285075129144428e-05, | |
| "loss": 9.8769, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 5.125, | |
| "learning_rate": 6.229051583116796e-05, | |
| "loss": 2.0862, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 4.875, | |
| "learning_rate": 6.173165676349103e-05, | |
| "loss": 2.2542, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 6.117419448671651e-05, | |
| "loss": 1.8006, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 6.0618149348164696e-05, | |
| "loss": 2.324, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 6.006354164343046e-05, | |
| "loss": 2.3711, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 5.9510391615642466e-05, | |
| "loss": 2.1159, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 5.8958719454724346e-05, | |
| "loss": 2.1294, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 37.25, | |
| "learning_rate": 5.840854529665767e-05, | |
| "loss": 1.9822, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 31.875, | |
| "learning_rate": 5.785988922274711e-05, | |
| "loss": 2.5124, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 5.375, | |
| "learning_rate": 5.7312771258887386e-05, | |
| "loss": 2.2204, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 5.676721137483225e-05, | |
| "loss": 2.1498, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 5.622322948346594e-05, | |
| "loss": 2.4613, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 5.568084544007588e-05, | |
| "loss": 2.8178, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 5.5140079041628214e-05, | |
| "loss": 2.6615, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 2960.0, | |
| "learning_rate": 5.4600950026045326e-05, | |
| "loss": 4.0852, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 6.0, | |
| "learning_rate": 5.406347807148515e-05, | |
| "loss": 2.5659, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 5.3527682795623146e-05, | |
| "loss": 2.181, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 5.2993583754936126e-05, | |
| "loss": 1.9782, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 5.246120044398839e-05, | |
| "loss": 2.0542, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 4.25, | |
| "learning_rate": 5.193055229472045e-05, | |
| "loss": 2.3066, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 5.14016586757394e-05, | |
| "loss": 2.1146, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 5.125, | |
| "learning_rate": 5.087453889161229e-05, | |
| "loss": 2.7877, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 5.0349212182161254e-05, | |
| "loss": 2.4127, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 4.98256977217614e-05, | |
| "loss": 2.1602, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 4.9304014618640995e-05, | |
| "loss": 1.8885, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 28.75, | |
| "learning_rate": 4.87841819141838e-05, | |
| "loss": 8.8349, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 4.826621858223431e-05, | |
| "loss": 1.738, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 5.25, | |
| "learning_rate": 4.7750143528405126e-05, | |
| "loss": 2.5058, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 4.723597558938672e-05, | |
| "loss": 2.0976, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 4.672373353226023e-05, | |
| "loss": 2.1503, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 4.6213436053812144e-05, | |
| "loss": 1.9976, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.5705101779852135e-05, | |
| "loss": 3.1302, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 27.875, | |
| "learning_rate": 4.519874926453302e-05, | |
| "loss": 8.4862, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 5.75, | |
| "learning_rate": 4.469439698967359e-05, | |
| "loss": 1.7939, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 13.125, | |
| "learning_rate": 4.419206336408418e-05, | |
| "loss": 2.2682, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 4.3691766722894435e-05, | |
| "loss": 2.3877, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 5.5, | |
| "learning_rate": 4.3193525326884435e-05, | |
| "loss": 2.703, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 4.26973573618179e-05, | |
| "loss": 2.3439, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 5.125, | |
| "learning_rate": 4.220328093777851e-05, | |
| "loss": 2.6788, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 4.1711314088509e-05, | |
| "loss": 2.2618, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 2.3582, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 4.0733780863598335e-05, | |
| "loss": 2.6398, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 2.7038376331329346, | |
| "eval_runtime": 1.7504, | |
| "eval_samples_per_second": 19.424, | |
| "eval_steps_per_second": 19.424, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 4.0248250167827275e-05, | |
| "loss": 2.1894, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 3.976490040526394e-05, | |
| "loss": 2.2171, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.5, | |
| "learning_rate": 3.9283749218128885e-05, | |
| "loss": 2.514, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 75.5, | |
| "learning_rate": 3.88048141683948e-05, | |
| "loss": 2.8149, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 3.832811273714569e-05, | |
| "loss": 2.3044, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 3.785366232393861e-05, | |
| "loss": 2.2342, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 3.738148024616863e-05, | |
| "loss": 2.2901, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 3.691158373843694e-05, | |
| "loss": 2.0687, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.644398995192147e-05, | |
| "loss": 2.7686, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 3.597871595375121e-05, | |
| "loss": 2.3497, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 3.5515778726382966e-05, | |
| "loss": 2.0393, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 41.0, | |
| "learning_rate": 3.5055195166981645e-05, | |
| "loss": 2.5633, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.375, | |
| "learning_rate": 3.459698208680359e-05, | |
| "loss": 2.3496, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 10.625, | |
| "learning_rate": 3.4141156210582756e-05, | |
| "loss": 4.4862, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 3.36877341759205e-05, | |
| "loss": 2.6964, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 3.3236732532678096e-05, | |
| "loss": 2.3155, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.5, | |
| "learning_rate": 3.2788167742372725e-05, | |
| "loss": 2.3611, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 3.234205617757686e-05, | |
| "loss": 1.846, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 3.1898414121320276e-05, | |
| "loss": 1.8453, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 107.0, | |
| "learning_rate": 3.1457257766496015e-05, | |
| "loss": 2.3292, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 5.90625, | |
| "learning_rate": 3.101860321526924e-05, | |
| "loss": 2.655, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 3.0582466478489455e-05, | |
| "loss": 2.2701, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.625, | |
| "learning_rate": 3.0148863475106314e-05, | |
| "loss": 2.1766, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 2.9717810031588277e-05, | |
| "loss": 2.3862, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 2.2309, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 2.886341466415412e-05, | |
| "loss": 2.2681, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 2.84401039255879e-05, | |
| "loss": 1.7952, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 2.8019405116448516e-05, | |
| "loss": 2.6097, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 2.7601333592202583e-05, | |
| "loss": 1.8333, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 2.7185904612421176e-05, | |
| "loss": 1.6016, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 2.677313334022268e-05, | |
| "loss": 2.4065, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 2.6363034841719392e-05, | |
| "loss": 2.3659, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.5, | |
| "learning_rate": 2.59556240854677e-05, | |
| "loss": 2.1504, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.375, | |
| "learning_rate": 2.5550915941921526e-05, | |
| "loss": 2.4407, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 2.514892518288988e-05, | |
| "loss": 1.938, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 2.4749666480997337e-05, | |
| "loss": 1.6936, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 2.4353154409148637e-05, | |
| "loss": 2.4835, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 4.25, | |
| "learning_rate": 2.3959403439996907e-05, | |
| "loss": 2.2736, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 54.75, | |
| "learning_rate": 2.356842794541516e-05, | |
| "loss": 1.9168, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 2.318024219597196e-05, | |
| "loss": 2.3022, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.2794860360410342e-05, | |
| "loss": 1.7, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.241229650513077e-05, | |
| "loss": 1.8489, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 2.2032564593677774e-05, | |
| "loss": 2.0878, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 4.0, | |
| "learning_rate": 2.165567848623009e-05, | |
| "loss": 2.169, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 2.1281651939094992e-05, | |
| "loss": 2.2711, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 2.0910498604205986e-05, | |
| "loss": 2.1268, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 2.0542232028624586e-05, | |
| "loss": 1.8266, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 6.25, | |
| "learning_rate": 2.0176865654045974e-05, | |
| "loss": 2.7551, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 1.981441281630816e-05, | |
| "loss": 2.5119, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 14.9375, | |
| "learning_rate": 1.94548867449054e-05, | |
| "loss": 2.2502, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 6.375, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 1.5596, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 1.8744667284469575e-05, | |
| "loss": 2.0902, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 1.8393999818379525e-05, | |
| "loss": 2.5071, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 5.09375, | |
| "learning_rate": 1.804631096356435e-05, | |
| "loss": 1.9177, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 1.7701613410634365e-05, | |
| "loss": 1.7456, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 1.735991974101756e-05, | |
| "loss": 3.7113, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 1.7021242426500493e-05, | |
| "loss": 1.8664, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 14.6875, | |
| "learning_rate": 1.6685593828773095e-05, | |
| "loss": 8.2793, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 1.6352986198977325e-05, | |
| "loss": 2.4292, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 1.6023431677260214e-05, | |
| "loss": 1.657, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 5.125, | |
| "learning_rate": 1.5696942292330576e-05, | |
| "loss": 2.4565, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 1.5373529961019974e-05, | |
| "loss": 2.3324, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 1.5053206487847914e-05, | |
| "loss": 2.8164, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 1.4735983564590783e-05, | |
| "loss": 2.2615, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 1.442187276985526e-05, | |
| "loss": 2.0414, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 6.59375, | |
| "learning_rate": 1.4110885568655564e-05, | |
| "loss": 2.8476, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 3.828125, | |
| "learning_rate": 1.3803033311995072e-05, | |
| "loss": 1.8397, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 7.25, | |
| "learning_rate": 1.3498327236452013e-05, | |
| "loss": 2.8825, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 1.3196778463769255e-05, | |
| "loss": 2.5735, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 1.2898398000448443e-05, | |
| "loss": 2.2819, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 1.260319673734821e-05, | |
| "loss": 1.9974, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 1.2311185449286677e-05, | |
| "loss": 1.935, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 1.2022374794648228e-05, | |
| "loss": 2.685, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 1.1736775314994342e-05, | |
| "loss": 1.77, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 1.1454397434679021e-05, | |
| "loss": 2.3842, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 1.1175251460468117e-05, | |
| "loss": 1.6934, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 1.0899347581163221e-05, | |
| "loss": 2.4064, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.0, | |
| "learning_rate": 1.062669586722983e-05, | |
| "loss": 2.1773, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 1.0357306270429624e-05, | |
| "loss": 2.1731, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.0, | |
| "learning_rate": 1.0091188623457415e-05, | |
| "loss": 2.1143, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 3.875, | |
| "learning_rate": 9.828352639582072e-06, | |
| "loss": 1.8268, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 9.568807912292077e-06, | |
| "loss": 2.2509, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 9.31256391494546e-06, | |
| "loss": 2.2418, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 5.5, | |
| "learning_rate": 9.05963000042378e-06, | |
| "loss": 2.7369, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 8.810015400790994e-06, | |
| "loss": 2.2038, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 8.563729226956319e-06, | |
| "loss": 2.3315, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 8.32078046834176e-06, | |
| "loss": 2.4199, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1096.0, | |
| "learning_rate": 8.081177992554013e-06, | |
| "loss": 2.5443, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 7.844930545060703e-06, | |
| "loss": 2.1696, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.0, | |
| "learning_rate": 7.612046748871327e-06, | |
| "loss": 1.9471, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.5, | |
| "learning_rate": 7.382535104222366e-06, | |
| "loss": 2.8064, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 7.156403988267069e-06, | |
| "loss": 2.7012, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 6.9336616547697965e-06, | |
| "loss": 2.7909, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 6.714316233804574e-06, | |
| "loss": 2.2768, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 6.498375731458528e-06, | |
| "loss": 2.1573, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.5, | |
| "learning_rate": 6.28584802953951e-06, | |
| "loss": 2.0637, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 6.076740885288479e-06, | |
| "loss": 2.5967, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.625, | |
| "learning_rate": 5.8710619310964445e-06, | |
| "loss": 2.5432, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 5.668818674225685e-06, | |
| "loss": 2.2381, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 18.625, | |
| "learning_rate": 5.470018496535967e-06, | |
| "loss": 2.029, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 5.274668654214932e-06, | |
| "loss": 2.1447, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 6.5, | |
| "learning_rate": 5.08277627751329e-06, | |
| "loss": 2.4182, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 2.0494, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 4.709391810729713e-06, | |
| "loss": 1.7768, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 4.527913349145441e-06, | |
| "loss": 2.0692, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 4.349919609678455e-06, | |
| "loss": 2.5842, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 4.175417089083378e-06, | |
| "loss": 2.354, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 26.875, | |
| "learning_rate": 4.004412156685711e-06, | |
| "loss": 2.6345, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 3.836911054149239e-06, | |
| "loss": 1.7726, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 3.6729198952483724e-06, | |
| "loss": 1.7874, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 3.512444665644865e-06, | |
| "loss": 2.2945, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 3.355491222669371e-06, | |
| "loss": 1.9434, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 3.202065295107726e-06, | |
| "loss": 2.1977, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 3.052172482991711e-06, | |
| "loss": 1.7724, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 2.905818257394799e-06, | |
| "loss": 2.1582, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 2.7630079602323442e-06, | |
| "loss": 1.8277, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 2.6237468040666512e-06, | |
| "loss": 2.0602, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 2.4880398719167586e-06, | |
| "loss": 2.0597, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 2.3558921170727888e-06, | |
| "loss": 2.1979, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 5.25, | |
| "learning_rate": 2.2273083629153147e-06, | |
| "loss": 1.8302, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.125, | |
| "learning_rate": 2.1022933027391555e-06, | |
| "loss": 2.534, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 1.9808514995821593e-06, | |
| "loss": 1.9609, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 1.8629873860586566e-06, | |
| "loss": 2.2896, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 1.7487052641976032e-06, | |
| "loss": 1.7238, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 3.84375, | |
| "learning_rate": 1.6380093052856483e-06, | |
| "loss": 2.0665, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.125, | |
| "learning_rate": 1.5309035497147684e-06, | |
| "loss": 2.7144, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 4.5, | |
| "learning_rate": 1.4273919068349184e-06, | |
| "loss": 2.633, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 8.125, | |
| "learning_rate": 1.3274781548112458e-06, | |
| "loss": 2.5471, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 1.231165940486234e-06, | |
| "loss": 2.6535, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 1.1384587792465872e-06, | |
| "loss": 2.3933, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.875, | |
| "learning_rate": 1.0493600548948878e-06, | |
| "loss": 2.4272, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 9.638730195261625e-07, | |
| "loss": 1.8742, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 8.820007934090879e-07, | |
| "loss": 2.5128, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 8.037463648721488e-07, | |
| "loss": 2.5476, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 7.291125901946027e-07, | |
| "loss": 2.1254, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.375, | |
| "learning_rate": 6.581021935021304e-07, | |
| "loss": 1.8256, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 5.907177666674812e-07, | |
| "loss": 2.4575, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 5.269617692158613e-07, | |
| "loss": 1.8101, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 4.668365282351372e-07, | |
| "loss": 2.6114, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 4.103442382909051e-07, | |
| "loss": 2.2765, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 3.5748696134639825e-07, | |
| "loss": 2.2232, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 3.0826662668720364e-07, | |
| "loss": 2.0377, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 6.0, | |
| "learning_rate": 2.6268503085089547e-07, | |
| "loss": 2.5245, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 2.2074383756137686e-07, | |
| "loss": 2.4988, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 1.824445776682504e-07, | |
| "loss": 2.7772, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 1.477886490908742e-07, | |
| "loss": 1.8278, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 1.1677731676733584e-07, | |
| "loss": 2.0586, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 8.941171260835601e-08, | |
| "loss": 1.8343, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 6.569283545587724e-08, | |
| "loss": 8.1851, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 4.562155104665955e-08, | |
| "loss": 2.291, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 3.875, | |
| "learning_rate": 2.9198591980705848e-08, | |
| "loss": 2.1164, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.375, | |
| "learning_rate": 1.642455769444995e-08, | |
| "loss": 2.2408, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 7.2999144389296335e-09, | |
| "loss": 2.0169, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 1.8249952627669154e-09, | |
| "loss": 2.4849, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 0.0, | |
| "loss": 1.7458, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.632516384124756, | |
| "eval_runtime": 1.7594, | |
| "eval_samples_per_second": 19.325, | |
| "eval_steps_per_second": 19.325, | |
| "step": 620 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 620, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 7881494668247040.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |