| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.877450980392156, | |
| "eval_steps": 500, | |
| "global_step": 408, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.875, | |
| "learning_rate": 2e-05, | |
| "loss": 7.5441, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 4e-05, | |
| "loss": 7.8502, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 5.0, | |
| "learning_rate": 6e-05, | |
| "loss": 7.9553, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.03125, | |
| "learning_rate": 8e-05, | |
| "loss": 7.4329, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 0.0001, | |
| "loss": 6.7046, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 0.00012, | |
| "loss": 6.1039, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 0.00014, | |
| "loss": 5.6611, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 0.00016, | |
| "loss": 5.0838, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 0.00018, | |
| "loss": 4.8258, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 0.0002, | |
| "loss": 4.4867, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 0.00019999688468941564, | |
| "loss": 4.0076, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.00019998753895176575, | |
| "loss": 3.8401, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.0001999719633693478, | |
| "loss": 3.8209, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 0.0001999501589126174, | |
| "loss": 3.7106, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 0.00019992212694012757, | |
| "loss": 3.7207, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00019988786919844436, | |
| "loss": 3.3611, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.0001998473878220379, | |
| "loss": 3.4981, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 0.00019980068533314934, | |
| "loss": 3.5695, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 0.00019974776464163387, | |
| "loss": 3.5129, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00019968862904477935, | |
| "loss": 3.4233, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.0001996232822271007, | |
| "loss": 3.4322, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00019955172826011062, | |
| "loss": 3.335, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0001994739716020657, | |
| "loss": 3.5036, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001993900170976888, | |
| "loss": 3.3962, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00019929986997786699, | |
| "loss": 3.3721, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00019920353585932578, | |
| "loss": 3.4649, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0001991010207442792, | |
| "loss": 3.2541, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00019899233102005573, | |
| "loss": 3.3746, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00019887747345870028, | |
| "loss": 3.3363, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.0001987564552165524, | |
| "loss": 3.3452, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0001986292838338003, | |
| "loss": 3.3879, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00019849596723401107, | |
| "loss": 3.2731, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 0.000198356513723637, | |
| "loss": 3.3342, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00019821093199149804, | |
| "loss": 3.3667, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0001980592311082404, | |
| "loss": 3.3652, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0001979014205257715, | |
| "loss": 3.2905, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019773751007667073, | |
| "loss": 3.3441, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0001975675099735774, | |
| "loss": 3.3104, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 0.00019739143080855378, | |
| "loss": 3.3328, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00019720928355242568, | |
| "loss": 3.2689, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.00019702107955409863, | |
| "loss": 3.2674, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 0.00019682683053985072, | |
| "loss": 3.246, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.0001966265486126022, | |
| "loss": 3.2299, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00019642024625116117, | |
| "loss": 3.347, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0001962079363094463, | |
| "loss": 3.2526, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.00019598963201568573, | |
| "loss": 3.1977, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.00019576534697159296, | |
| "loss": 3.3306, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 0.0001955350951515195, | |
| "loss": 3.2819, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00019529889090158392, | |
| "loss": 3.2065, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 0.0001950567489387783, | |
| "loss": 3.2446, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00019480868435005095, | |
| "loss": 3.1567, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001945547125913667, | |
| "loss": 3.2422, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00019429484948674372, | |
| "loss": 3.2118, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00019402911122726757, | |
| "loss": 3.1756, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00019375751437008252, | |
| "loss": 3.1951, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00019348007583735983, | |
| "loss": 3.1604, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.0001931968129152435, | |
| "loss": 3.1231, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019290774325277305, | |
| "loss": 3.1371, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00019261288486078414, | |
| "loss": 3.2218, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.0001923122561107861, | |
| "loss": 3.178, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00019200587573381744, | |
| "loss": 3.1543, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00019169376281927888, | |
| "loss": 3.1088, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 0.0001913759368137437, | |
| "loss": 3.1465, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00019105241751974622, | |
| "loss": 3.1484, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00019072322509454815, | |
| "loss": 3.127, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.0001903883800488824, | |
| "loss": 3.123, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00019004790324567519, | |
| "loss": 3.1802, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00018970181589874637, | |
| "loss": 3.0633, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018935013957148742, | |
| "loss": 3.1452, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00018899289617551804, | |
| "loss": 3.0458, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0001886301079693209, | |
| "loss": 3.1749, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 0.0001882617975568547, | |
| "loss": 3.1142, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00018788798788614607, | |
| "loss": 3.0497, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00018750870224785939, | |
| "loss": 3.1338, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00018712396427384594, | |
| "loss": 3.1228, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00018673379793567146, | |
| "loss": 3.1073, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 0.00018633822754312234, | |
| "loss": 3.0681, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.0001859372777426912, | |
| "loss": 3.0805, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00018553097351604118, | |
| "loss": 3.0641, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00018511934017844948, | |
| "loss": 3.011, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00018470240337722991, | |
| "loss": 3.0556, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00018428018909013506, | |
| "loss": 3.1694, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00018385272362373775, | |
| "loss": 3.1034, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00018342003361179176, | |
| "loss": 3.0954, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0001829821460135726, | |
| "loss": 3.0729, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00018253908811219764, | |
| "loss": 3.0459, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00018209088751292626, | |
| "loss": 3.0627, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 0.00018163757214143992, | |
| "loss": 3.0817, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00018117917024210208, | |
| "loss": 3.0809, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00018071571037619853, | |
| "loss": 3.0377, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00018024722142015781, | |
| "loss": 3.0733, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00017977373256375194, | |
| "loss": 3.0641, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00017929527330827786, | |
| "loss": 3.1211, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00017881187346471925, | |
| "loss": 3.1069, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00017832356315188906, | |
| "loss": 2.9687, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00017783037279455298, | |
| "loss": 3.1262, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00017733233312153393, | |
| "loss": 3.0263, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 0.00017682947516379707, | |
| "loss": 3.0806, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00017632183025251686, | |
| "loss": 3.0914, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00017580943001712455, | |
| "loss": 3.0555, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00017529230638333772, | |
| "loss": 2.9838, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00017477049157117093, | |
| "loss": 2.9647, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00017424401809292833, | |
| "loss": 3.03, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.0001737129187511779, | |
| "loss": 3.0758, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.0001731772266367077, | |
| "loss": 3.035, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00017263697512646394, | |
| "loss": 3.0058, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00017209219788147167, | |
| "loss": 2.9659, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00017154292884473713, | |
| "loss": 2.8561, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001709892022391333, | |
| "loss": 2.9509, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00017043105256526724, | |
| "loss": 2.998, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00016986851459933067, | |
| "loss": 2.9808, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00016930162339093318, | |
| "loss": 2.9386, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00016873041426091845, | |
| "loss": 2.9435, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.0001681549227991634, | |
| "loss": 2.9056, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00016757518486236087, | |
| "loss": 2.9288, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00016699123657178553, | |
| "loss": 2.9522, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.0001664031143110431, | |
| "loss": 2.838, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00016581085472380376, | |
| "loss": 2.9151, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00016521449471151867, | |
| "loss": 2.9665, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00016461407143112097, | |
| "loss": 2.9064, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00016400962229271072, | |
| "loss": 2.8611, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00016340118495722388, | |
| "loss": 2.9698, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00016278879733408585, | |
| "loss": 2.8706, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00016217249757884955, | |
| "loss": 3.0017, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00016155232409081793, | |
| "loss": 2.9088, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001609283155106517, | |
| "loss": 2.8959, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00016030051071796146, | |
| "loss": 2.909, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00015966894882888562, | |
| "loss": 2.9291, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00015903366919365282, | |
| "loss": 2.9638, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00015839471139413066, | |
| "loss": 2.9725, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.0001577521152413589, | |
| "loss": 3.0265, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.0001571059207730695, | |
| "loss": 2.8662, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001564561682511918, | |
| "loss": 2.9864, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00015580289815934401, | |
| "loss": 2.9896, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00015514615120031076, | |
| "loss": 2.9797, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00015448596829350706, | |
| "loss": 2.9245, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00015382239057242888, | |
| "loss": 2.9472, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00015315545938209015, | |
| "loss": 3.0054, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00015248521627644684, | |
| "loss": 3.002, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00015181170301580777, | |
| "loss": 2.9129, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0001511349615642327, | |
| "loss": 2.9397, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00015045503408691775, | |
| "loss": 2.9214, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00014977196294756832, | |
| "loss": 2.9527, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00014908579070575936, | |
| "loss": 2.9482, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00014839656011428389, | |
| "loss": 3.0022, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00014770431411648897, | |
| "loss": 2.9553, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.0001470090958436003, | |
| "loss": 2.9322, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.0001463109486120348, | |
| "loss": 2.9479, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00014560991592070158, | |
| "loss": 2.8904, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00014490604144829202, | |
| "loss": 2.9523, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00014419936905055793, | |
| "loss": 2.9384, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00014348994275757931, | |
| "loss": 2.9068, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00014277780677102097, | |
| "loss": 2.9824, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00014206300546137842, | |
| "loss": 2.9305, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00014134558336521342, | |
| "loss": 2.9497, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00014062558518237892, | |
| "loss": 2.9443, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.0001399030557732341, | |
| "loss": 2.7934, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00013917804015584932, | |
| "loss": 2.8409, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00013845058350320108, | |
| "loss": 2.7915, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00013772073114035762, | |
| "loss": 2.8044, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.0001369885285416547, | |
| "loss": 2.7569, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00013625402132786248, | |
| "loss": 2.7233, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00013551725526334284, | |
| "loss": 2.8137, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00013477827625319824, | |
| "loss": 2.7915, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.0001340371303404113, | |
| "loss": 2.8062, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00013329386370297615, | |
| "loss": 2.7944, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00013254852265102117, | |
| "loss": 2.832, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00013180115362392382, | |
| "loss": 2.802, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.0001310518031874169, | |
| "loss": 2.7101, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00013030051803068727, | |
| "loss": 2.8507, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00012954734496346704, | |
| "loss": 2.8022, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00012879233091311667, | |
| "loss": 2.8003, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00012803552292170144, | |
| "loss": 2.8, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00012727696814306033, | |
| "loss": 2.9019, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00012651671383986788, | |
| "loss": 2.815, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 0.0001257548073806897, | |
| "loss": 2.8604, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00012499129623703086, | |
| "loss": 2.7868, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00012422622798037832, | |
| "loss": 2.8025, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.0001234596502792369, | |
| "loss": 2.7852, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.000122691610896159, | |
| "loss": 2.7422, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.00012192215768476916, | |
| "loss": 2.7908, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00012115133858678191, | |
| "loss": 2.8865, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00012037920162901521, | |
| "loss": 2.8135, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00011960579492039783, | |
| "loss": 2.8735, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00011883116664897178, | |
| "loss": 2.7555, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00011805536507889021, | |
| "loss": 2.7936, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00011727843854740996, | |
| "loss": 2.8138, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00011650043546187995, | |
| "loss": 2.8566, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00011572140429672508, | |
| "loss": 2.8295, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.0001149413935904261, | |
| "loss": 2.8278, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00011416045194249516, | |
| "loss": 2.7748, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00011337862801044792, | |
| "loss": 2.838, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 0.00011259597050677178, | |
| "loss": 2.7914, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00011181252819589081, | |
| "loss": 2.7811, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00011102834989112751, | |
| "loss": 2.8006, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00011024348445166133, | |
| "loss": 2.8152, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.0001094579807794845, | |
| "loss": 2.8045, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00010867188781635512, | |
| "loss": 2.8115, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00010788525454074765, | |
| "loss": 2.8645, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.0001070981299648016, | |
| "loss": 2.8005, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00010631056313126734, | |
| "loss": 2.8381, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00010552260311045082, | |
| "loss": 2.8413, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00010473429899715581, | |
| "loss": 2.8571, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00010394569990762529, | |
| "loss": 2.8747, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.00010315685497648106, | |
| "loss": 2.8742, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00010236781335366239, | |
| "loss": 2.8503, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.0001015786242013637, | |
| "loss": 2.8812, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00010078933669097135, | |
| "loss": 2.8629, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.0001, | |
| "loss": 2.7461, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 9.92106633090287e-05, | |
| "loss": 2.6341, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 9.842137579863632e-05, | |
| "loss": 2.7251, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 9.763218664633763e-05, | |
| "loss": 2.6282, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 9.684314502351894e-05, | |
| "loss": 2.6809, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 9.605430009237474e-05, | |
| "loss": 2.6245, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 9.526570100284422e-05, | |
| "loss": 2.6561, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 9.447739688954919e-05, | |
| "loss": 2.7119, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 9.368943686873267e-05, | |
| "loss": 2.7143, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 9.29018700351984e-05, | |
| "loss": 2.7154, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 9.211474545925236e-05, | |
| "loss": 2.7656, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 9.132811218364495e-05, | |
| "loss": 2.6825, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 9.054201922051552e-05, | |
| "loss": 2.7077, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 8.975651554833869e-05, | |
| "loss": 2.6552, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 8.89716501088725e-05, | |
| "loss": 2.6438, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 8.818747180410921e-05, | |
| "loss": 2.685, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 8.740402949322827e-05, | |
| "loss": 2.6834, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 8.66213719895521e-05, | |
| "loss": 2.6793, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 8.583954805750487e-05, | |
| "loss": 2.6911, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 8.505860640957391e-05, | |
| "loss": 2.6713, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 8.427859570327494e-05, | |
| "loss": 2.6732, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 0.5, | |
| "learning_rate": 8.349956453812009e-05, | |
| "loss": 2.7043, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 8.272156145259006e-05, | |
| "loss": 2.6899, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "grad_norm": 0.5, | |
| "learning_rate": 8.194463492110981e-05, | |
| "loss": 2.6345, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 8.11688333510282e-05, | |
| "loss": 2.7625, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 8.03942050796022e-05, | |
| "loss": 2.676, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 7.96207983709848e-05, | |
| "loss": 2.6027, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 7.88486614132181e-05, | |
| "loss": 2.6139, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 7.807784231523089e-05, | |
| "loss": 2.7281, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 7.730838910384097e-05, | |
| "loss": 2.7225, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 7.654034972076314e-05, | |
| "loss": 2.6199, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 0.5, | |
| "learning_rate": 7.57737720196217e-05, | |
| "loss": 2.721, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 7.500870376296918e-05, | |
| "loss": 2.6753, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 7.424519261931036e-05, | |
| "loss": 2.6821, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 7.348328616013213e-05, | |
| "loss": 2.6376, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 7.27230318569397e-05, | |
| "loss": 2.8049, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 7.196447707829857e-05, | |
| "loss": 2.6317, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 7.120766908688336e-05, | |
| "loss": 2.6795, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 0.5, | |
| "learning_rate": 7.045265503653303e-05, | |
| "loss": 2.752, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 6.969948196931272e-05, | |
| "loss": 2.6873, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 6.894819681258312e-05, | |
| "loss": 2.5891, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 6.819884637607619e-05, | |
| "loss": 2.6957, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 6.745147734897883e-05, | |
| "loss": 2.6414, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 6.670613629702391e-05, | |
| "loss": 2.6997, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 6.596286965958872e-05, | |
| "loss": 2.6878, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 6.522172374680177e-05, | |
| "loss": 2.7198, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 6.448274473665717e-05, | |
| "loss": 2.7487, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 0.5, | |
| "learning_rate": 6.374597867213756e-05, | |
| "loss": 2.7132, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 6.301147145834534e-05, | |
| "loss": 2.7664, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 6.22792688596424e-05, | |
| "loss": 2.7209, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 6.154941649679894e-05, | |
| "loss": 2.7295, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 0.5, | |
| "learning_rate": 6.0821959844150687e-05, | |
| "loss": 2.7088, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 6.00969442267659e-05, | |
| "loss": 2.5827, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 5.9374414817621114e-05, | |
| "loss": 2.5989, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 5.8654416634786605e-05, | |
| "loss": 2.6026, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 5.7936994538621605e-05, | |
| "loss": 2.5145, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 5.7222193228979037e-05, | |
| "loss": 2.464, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 5.651005724242071e-05, | |
| "loss": 2.5677, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 5.58006309494421e-05, | |
| "loss": 2.5876, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 5.16, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 5.509395855170798e-05, | |
| "loss": 2.566, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 5.43900840792984e-05, | |
| "loss": 2.5722, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 5.368905138796523e-05, | |
| "loss": 2.5799, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.22, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 5.2990904156399726e-05, | |
| "loss": 2.5271, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 5.24, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 5.229568588351108e-05, | |
| "loss": 2.5608, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 5.25, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 5.160343988571613e-05, | |
| "loss": 2.4864, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 5.0914209294240644e-05, | |
| "loss": 2.6039, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 5.022803705243169e-05, | |
| "loss": 2.6246, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 4.9544965913082264e-05, | |
| "loss": 2.5739, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 4.886503843576735e-05, | |
| "loss": 2.6554, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 4.818829698419225e-05, | |
| "loss": 2.5539, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 4.751478372355317e-05, | |
| "loss": 2.512, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 5.39, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 4.684454061790987e-05, | |
| "loss": 2.5418, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 4.617760942757117e-05, | |
| "loss": 2.5095, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 4.5514031706492986e-05, | |
| "loss": 2.5265, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 4.485384879968926e-05, | |
| "loss": 2.5866, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 5.47, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 4.4197101840655995e-05, | |
| "loss": 2.5831, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 4.354383174880818e-05, | |
| "loss": 2.5747, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "grad_norm": 0.625, | |
| "learning_rate": 4.289407922693053e-05, | |
| "loss": 2.6402, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 5.53, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 4.224788475864115e-05, | |
| "loss": 2.6473, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 5.55, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 4.1605288605869365e-05, | |
| "loss": 2.4913, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 4.0966330806347166e-05, | |
| "loss": 2.5025, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 5.59, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 4.033105117111441e-05, | |
| "loss": 2.5332, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 3.969948928203856e-05, | |
| "loss": 2.5641, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 3.907168448934836e-05, | |
| "loss": 2.5393, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "grad_norm": 0.625, | |
| "learning_rate": 3.844767590918209e-05, | |
| "loss": 2.5763, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 5.67, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 3.7827502421150496e-05, | |
| "loss": 2.5888, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 5.69, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 3.7211202665914155e-05, | |
| "loss": 2.6156, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 3.659881504277613e-05, | |
| "loss": 2.5849, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 3.599037770728929e-05, | |
| "loss": 2.5871, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 3.538592856887901e-05, | |
| "loss": 2.662, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 3.478550528848134e-05, | |
| "loss": 2.6323, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 5.78, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 3.4189145276196245e-05, | |
| "loss": 2.6352, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 3.359688568895689e-05, | |
| "loss": 2.5697, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 5.82, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 3.3008763428214505e-05, | |
| "loss": 2.5935, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 3.242481513763913e-05, | |
| "loss": 2.5634, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 3.1845077200836636e-05, | |
| "loss": 2.5948, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 3.126958573908156e-05, | |
| "loss": 2.5673, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 3.0698376609066825e-05, | |
| "loss": 2.5718, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 3.0131485400669356e-05, | |
| "loss": 2.5712, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 2.9568947434732775e-05, | |
| "loss": 2.5854, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 5.96, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 2.9010797760866737e-05, | |
| "loss": 2.5602, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 2.8457071155262884e-05, | |
| "loss": 2.587, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 2.7907802118528383e-05, | |
| "loss": 2.608, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 2.736302487353609e-05, | |
| "loss": 2.5485, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 2.682277336329233e-05, | |
| "loss": 2.498, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 2.628708124882212e-05, | |
| "loss": 2.5104, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 6.06, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 2.575598190707168e-05, | |
| "loss": 2.4751, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 2.5229508428829096e-05, | |
| "loss": 2.512, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 2.4707693616662308e-05, | |
| "loss": 2.4899, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 2.4190569982875467e-05, | |
| "loss": 2.4877, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 6.14, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 2.367816974748317e-05, | |
| "loss": 2.482, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 6.16, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 2.3170524836202933e-05, | |
| "loss": 2.4958, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.18, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 2.266766687846611e-05, | |
| "loss": 2.4936, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 2.216962720544703e-05, | |
| "loss": 2.5029, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 6.22, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 2.167643684811096e-05, | |
| "loss": 2.4946, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 6.24, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 2.1188126535280773e-05, | |
| "loss": 2.4766, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 2.070472669172213e-05, | |
| "loss": 2.4435, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.28, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 2.022626743624807e-05, | |
| "loss": 2.5658, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 1.9752778579842213e-05, | |
| "loss": 2.514, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 6.32, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 1.9284289623801477e-05, | |
| "loss": 2.4731, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 6.34, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 1.882082975789795e-05, | |
| "loss": 2.4501, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 1.8362427858560093e-05, | |
| "loss": 2.531, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 6.38, | |
| "grad_norm": 0.625, | |
| "learning_rate": 1.7909112487073754e-05, | |
| "loss": 2.5093, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 1.74609118878024e-05, | |
| "loss": 2.4216, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 6.42, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 1.7017853986427425e-05, | |
| "loss": 2.5315, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 6.44, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 1.657996638820826e-05, | |
| "loss": 2.5197, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 1.6147276376262255e-05, | |
| "loss": 2.3946, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 6.48, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 1.5719810909864942e-05, | |
| "loss": 2.5408, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 1.5297596622770115e-05, | |
| "loss": 2.4942, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 6.51, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 1.4880659821550546e-05, | |
| "loss": 2.434, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 6.53, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 1.4469026483958837e-05, | |
| "loss": 2.5376, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 6.55, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 1.4062722257308803e-05, | |
| "loss": 2.5387, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 1.3661772456877675e-05, | |
| "loss": 2.4496, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 6.59, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 1.3266202064328548e-05, | |
| "loss": 2.5007, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 1.2876035726154045e-05, | |
| "loss": 2.4802, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 6.63, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 1.2491297752140641e-05, | |
| "loss": 2.5287, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 1.2112012113853954e-05, | |
| "loss": 2.4877, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 1.1738202443145308e-05, | |
| "loss": 2.4865, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 6.69, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 1.1369892030679141e-05, | |
| "loss": 2.4497, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 6.71, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 1.1007103824481979e-05, | |
| "loss": 2.486, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 1.0649860428512604e-05, | |
| "loss": 2.5043, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 6.75, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 1.029818410125365e-05, | |
| "loss": 2.4816, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.77, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 9.952096754324847e-06, | |
| "loss": 2.5321, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 9.611619951117657e-06, | |
| "loss": 2.5113, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 6.81, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 9.276774905451869e-06, | |
| "loss": 2.4642, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 6.83, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 8.94758248025378e-06, | |
| "loss": 2.5338, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 8.624063186256326e-06, | |
| "loss": 2.4683, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 6.87, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 8.306237180721121e-06, | |
| "loss": 2.4918, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 7.994124266182568e-06, | |
| "loss": 2.4688, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 6.91, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 7.687743889213938e-06, | |
| "loss": 2.4914, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 6.93, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 7.387115139215894e-06, | |
| "loss": 2.4586, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 6.95, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 7.0922567472269444e-06, | |
| "loss": 2.4991, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "grad_norm": 0.625, | |
| "learning_rate": 6.803187084756524e-06, | |
| "loss": 2.5431, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 6.519924162640167e-06, | |
| "loss": 2.4695, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.625, | |
| "learning_rate": 6.242485629917494e-06, | |
| "loss": 2.475, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 5.9708887727324525e-06, | |
| "loss": 2.4779, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 5.7051505132562965e-06, | |
| "loss": 2.5005, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 5.445287408633304e-06, | |
| "loss": 2.4736, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 5.191315649949047e-06, | |
| "loss": 2.4719, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 7.09, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 4.943251061221721e-06, | |
| "loss": 2.5049, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 4.701109098416079e-06, | |
| "loss": 2.507, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 7.13, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 4.464904848480523e-06, | |
| "loss": 2.4379, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 7.15, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 4.234653028407054e-06, | |
| "loss": 2.4399, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 7.17, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 4.0103679843142895e-06, | |
| "loss": 2.4656, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 3.7920636905537155e-06, | |
| "loss": 2.493, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 3.5797537488388323e-06, | |
| "loss": 2.4273, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 3.373451387397819e-06, | |
| "loss": 2.4495, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.25, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 3.1731694601492833e-06, | |
| "loss": 2.4133, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 7.27, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 2.9789204459013785e-06, | |
| "loss": 2.4453, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 2.7907164475743043e-06, | |
| "loss": 2.4051, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 2.6085691914462306e-06, | |
| "loss": 2.4648, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 2.4324900264226403e-06, | |
| "loss": 2.5328, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 7.35, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 2.2624899233292806e-06, | |
| "loss": 2.4942, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 2.098579474228546e-06, | |
| "loss": 2.402, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 7.39, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 1.9407688917595925e-06, | |
| "loss": 2.4559, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 7.41, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 1.7890680085019595e-06, | |
| "loss": 2.47, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 7.43, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 1.6434862763630155e-06, | |
| "loss": 2.5025, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 7.45, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 1.5040327659889608e-06, | |
| "loss": 2.4876, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 1.370716166199726e-06, | |
| "loss": 2.4821, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 7.49, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 1.2435447834476255e-06, | |
| "loss": 2.5129, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 1.122526541299751e-06, | |
| "loss": 2.5025, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 7.52, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 1.0076689799442873e-06, | |
| "loss": 2.4836, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 7.54, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 8.989792557207887e-07, | |
| "loss": 2.5049, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 7.56, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 7.964641406742135e-07, | |
| "loss": 2.4673, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 7.58, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 7.001300221330387e-07, | |
| "loss": 2.4203, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 6.099829023112235e-07, | |
| "loss": 2.4317, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 7.62, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 5.260283979343084e-07, | |
| "loss": 2.4273, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 4.482717398894165e-07, | |
| "loss": 2.539, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 3.767177728993265e-07, | |
| "loss": 2.4855, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 7.68, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 3.1137095522068007e-07, | |
| "loss": 2.525, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 2.522353583661263e-07, | |
| "loss": 2.3878, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 1.9931466685065847e-07, | |
| "loss": 2.4339, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.74, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 1.5261217796211923e-07, | |
| "loss": 2.4491, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 1.1213080155564326e-07, | |
| "loss": 2.4598, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 7.787305987243532e-08, | |
| "loss": 2.4383, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 4.9841087382618276e-08, | |
| "loss": 2.4653, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 7.82, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 2.8036630652206187e-08, | |
| "loss": 2.4291, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 7.84, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 1.2461048234269079e-08, | |
| "loss": 2.5025, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "grad_norm": 0.625, | |
| "learning_rate": 3.115310584367315e-09, | |
| "loss": 2.4367, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 7.88, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.0, | |
| "loss": 2.5149, | |
| "step": 408 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 408, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 51, | |
| "total_flos": 6.078120159011144e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |