| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 376, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02127659574468085, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 2e-05, | |
| "loss": 3.872, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0425531914893617, | |
| "grad_norm": 0.83203125, | |
| "learning_rate": 4e-05, | |
| "loss": 3.9714, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06382978723404255, | |
| "grad_norm": 0.82421875, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9503, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 8e-05, | |
| "loss": 4.0784, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.10638297872340426, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 0.0001, | |
| "loss": 3.9539, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1276595744680851, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.00012, | |
| "loss": 3.8024, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14893617021276595, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.00014, | |
| "loss": 3.6005, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 0.00016, | |
| "loss": 3.8633, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.19148936170212766, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.00018, | |
| "loss": 3.6551, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2127659574468085, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.0002, | |
| "loss": 3.6645, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.23404255319148937, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.000199996316124771, | |
| "loss": 3.7208, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2553191489361702, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.0001999852647705027, | |
| "loss": 3.6193, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2765957446808511, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001999668467514313, | |
| "loss": 3.7057, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2978723404255319, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00019994106342455053, | |
| "loss": 3.5714, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.3191489361702128, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.00019990791668951155, | |
| "loss": 3.582, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00019986740898848306, | |
| "loss": 3.5228, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3617021276595745, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00019981954330597143, | |
| "loss": 3.5893, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3829787234042553, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00019976432316860067, | |
| "loss": 3.5203, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.40425531914893614, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 0.00019970175264485266, | |
| "loss": 3.5939, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 0.00019963183634476756, | |
| "loss": 3.5296, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.44680851063829785, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.00019955457941960383, | |
| "loss": 3.6242, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.46808510638297873, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 0.0001994699875614589, | |
| "loss": 3.517, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.48936170212765956, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00019937806700284986, | |
| "loss": 3.5748, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.00019927882451625402, | |
| "loss": 3.5263, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5319148936170213, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00019917226741361015, | |
| "loss": 3.5638, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5531914893617021, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00019905840354577972, | |
| "loss": 3.5424, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.574468085106383, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 0.00019893724130196828, | |
| "loss": 3.5726, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5957446808510638, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 0.00019880878960910772, | |
| "loss": 3.5656, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.6170212765957447, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 0.00019867305793119816, | |
| "loss": 3.6008, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6382978723404256, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 0.0001985300562686109, | |
| "loss": 3.5136, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6595744680851063, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 0.00019837979515735166, | |
| "loss": 3.5632, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.0001982222856682841, | |
| "loss": 3.6284, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.7021276595744681, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001980575394063143, | |
| "loss": 3.4885, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.723404255319149, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.0001978855685095358, | |
| "loss": 3.6102, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7446808510638298, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0001977063856483351, | |
| "loss": 3.5844, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7659574468085106, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00019752000402445825, | |
| "loss": 3.5097, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7872340425531915, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00019732643737003827, | |
| "loss": 3.492, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.8085106382978723, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00019712569994658315, | |
| "loss": 3.6192, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.8297872340425532, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.00019691780654392535, | |
| "loss": 3.6314, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00019670277247913205, | |
| "loss": 3.6429, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8723404255319149, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00019648061359537646, | |
| "loss": 3.5714, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8936170212765957, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019625134626077083, | |
| "loss": 3.574, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.9148936170212766, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 0.00019601498736716017, | |
| "loss": 3.6269, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.9361702127659575, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00019577155432887804, | |
| "loss": 3.659, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9574468085106383, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00019552106508146318, | |
| "loss": 3.6223, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9787234042553191, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00019526353808033825, | |
| "loss": 3.6389, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00019499899229945012, | |
| "loss": 3.4551, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.0212765957446808, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.0001947274472298717, | |
| "loss": 3.2727, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.0425531914893618, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00019444892287836613, | |
| "loss": 3.3136, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.00019416343976591261, | |
| "loss": 3.3188, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0851063829787233, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 0.00019387101892619443, | |
| "loss": 3.424, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.1063829787234043, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00019357168190404936, | |
| "loss": 3.3676, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.127659574468085, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.00019326545075388225, | |
| "loss": 3.3535, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.148936170212766, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.00019295234803804004, | |
| "loss": 3.1686, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.1702127659574468, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00019263239682514952, | |
| "loss": 3.3986, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.1914893617021276, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.0001923056206884176, | |
| "loss": 3.2916, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.2127659574468086, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00019197204370389467, | |
| "loss": 3.3444, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.2340425531914894, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001916316904487005, | |
| "loss": 3.3603, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.2553191489361701, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00019128458599921357, | |
| "loss": 3.308, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 0.00019093075592922358, | |
| "loss": 3.4154, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.297872340425532, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00019057022630804716, | |
| "loss": 3.3095, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.3191489361702127, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.00019020302369860708, | |
| "loss": 3.3266, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.3404255319148937, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.0001898291751554753, | |
| "loss": 3.276, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.3617021276595744, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00018944870822287956, | |
| "loss": 3.3428, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.3829787234042552, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00018906165093267405, | |
| "loss": 3.2515, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.4042553191489362, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00018866803180227402, | |
| "loss": 3.3125, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.425531914893617, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00018826787983255473, | |
| "loss": 3.2968, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.4468085106382977, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00018786122450571485, | |
| "loss": 3.3705, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.4680851063829787, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.00018744809578310397, | |
| "loss": 3.2878, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.4893617021276595, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00018702852410301554, | |
| "loss": 3.3214, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.5106382978723403, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 3.2708, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.5319148936170213, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.00018617017599480682, | |
| "loss": 3.3087, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.5531914893617023, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00018573146280763324, | |
| "loss": 3.3227, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.574468085106383, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.000185286433140216, | |
| "loss": 3.3296, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.5957446808510638, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001848351197812304, | |
| "loss": 3.3282, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.6170212765957448, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 0.00018437755598231856, | |
| "loss": 3.3421, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.6382978723404256, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00018391377545563938, | |
| "loss": 3.3002, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.6595744680851063, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00018344381237138472, | |
| "loss": 3.3293, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.6808510638297873, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 0.0001829677013552619, | |
| "loss": 3.3688, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 0.00018248547748594244, | |
| "loss": 3.2586, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.7234042553191489, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00018199717629247773, | |
| "loss": 3.3783, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.7446808510638299, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00018150283375168114, | |
| "loss": 3.3503, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.7659574468085106, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0001810024862854775, | |
| "loss": 3.2862, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.7872340425531914, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00018049617075821962, | |
| "loss": 3.2503, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.8085106382978724, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00017998392447397197, | |
| "loss": 3.3987, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.8297872340425532, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001794657851737625, | |
| "loss": 3.3948, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.851063829787234, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00017894179103280198, | |
| "loss": 3.414, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.872340425531915, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00017841198065767107, | |
| "loss": 3.3495, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.8936170212765957, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00017787639308347608, | |
| "loss": 3.3357, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.9148936170212765, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.000177335067770973, | |
| "loss": 3.3956, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.9361702127659575, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00017678804460366, | |
| "loss": 3.4261, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.9574468085106385, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00017623536388483905, | |
| "loss": 3.3929, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.978723404255319, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00017567706633464628, | |
| "loss": 3.4055, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.00017511319308705198, | |
| "loss": 3.0576, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.021276595744681, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 0.00017454378568683003, | |
| "loss": 3.1095, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0425531914893615, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 0.0001739688860864967, | |
| "loss": 3.1669, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0638297872340425, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00017338853664321992, | |
| "loss": 3.1293, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.0851063829787235, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.00017280278011569847, | |
| "loss": 3.2461, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.106382978723404, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00017221165966101163, | |
| "loss": 3.2222, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00017161521883143934, | |
| "loss": 3.1956, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.148936170212766, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.0001710135015712536, | |
| "loss": 3.0099, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.1702127659574466, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00017040655221348057, | |
| "loss": 3.2425, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.1914893617021276, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00016979441547663435, | |
| "loss": 3.1365, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.2127659574468086, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.00016917713646142222, | |
| "loss": 3.1903, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.2340425531914896, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00016855476064742155, | |
| "loss": 3.1938, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.25531914893617, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00016792733388972932, | |
| "loss": 3.1561, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.276595744680851, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.0001672949024155833, | |
| "loss": 3.259, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.297872340425532, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 0.00016665751282095634, | |
| "loss": 3.1575, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.3191489361702127, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00016601521206712318, | |
| "loss": 3.1849, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.3404255319148937, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.0001653680474772006, | |
| "loss": 3.1254, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.3617021276595747, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00016471606673266066, | |
| "loss": 3.1994, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.382978723404255, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00016405931786981755, | |
| "loss": 3.101, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.404255319148936, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00016339784927628867, | |
| "loss": 3.1611, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.425531914893617, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.0001627317096874294, | |
| "loss": 3.1622, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.4468085106382977, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00016206094818274229, | |
| "loss": 3.2131, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.4680851063829787, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001613856141822612, | |
| "loss": 3.1511, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.4893617021276597, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.00016070575744291004, | |
| "loss": 3.1662, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.5106382978723403, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00016002142805483685, | |
| "loss": 3.1092, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.5319148936170213, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.0001593326764377232, | |
| "loss": 3.1444, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00015863955333706957, | |
| "loss": 3.1738, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.574468085106383, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00015794210982045636, | |
| "loss": 3.1766, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.595744680851064, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00015724039727378148, | |
| "loss": 3.166, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.617021276595745, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00015653446739747427, | |
| "loss": 3.1837, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.6382978723404253, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00015582437220268647, | |
| "loss": 3.1519, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.6595744680851063, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00015511016400746, | |
| "loss": 3.165, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6808510638297873, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00015439189543287247, | |
| "loss": 3.2062, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.702127659574468, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.00015366961939916008, | |
| "loss": 3.0979, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.723404255319149, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.0001529433891218185, | |
| "loss": 3.217, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.74468085106383, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.0001522132581076825, | |
| "loss": 3.1789, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.7659574468085104, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.0001514792801509831, | |
| "loss": 3.1253, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.7872340425531914, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00015074150932938455, | |
| "loss": 3.0813, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.8085106382978724, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 3.2233, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.829787234042553, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00014925480679538647, | |
| "loss": 3.2241, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.851063829787234, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00014850598461951963, | |
| "loss": 3.2428, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.872340425531915, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00014775358864374885, | |
| "loss": 3.1833, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.8936170212765955, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.000146997674302732, | |
| "loss": 3.162, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.9148936170212765, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.0001462382972903515, | |
| "loss": 3.2095, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.9361702127659575, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.0001454755135556106, | |
| "loss": 3.2355, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.9574468085106385, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.0001447093792985114, | |
| "loss": 3.1969, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00014393995096591416, | |
| "loss": 3.2092, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 0.0001431672852473784, | |
| "loss": 2.7442, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 3.021276595744681, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 0.0001423914390709861, | |
| "loss": 2.9817, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 3.0425531914893615, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 0.00014161246959914744, | |
| "loss": 3.0423, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 3.0638297872340425, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00014083043422438935, | |
| "loss": 2.9844, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.0851063829787235, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 0.00014004539056512667, | |
| "loss": 3.0951, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.106382978723404, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 0.0001392573964614172, | |
| "loss": 3.089, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 3.127659574468085, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00013846650997070012, | |
| "loss": 3.0649, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 3.148936170212766, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00013767278936351854, | |
| "loss": 2.8683, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 3.1702127659574466, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 0.00013687629311922602, | |
| "loss": 3.1071, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 3.1914893617021276, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00013607707992167834, | |
| "loss": 3.0015, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.2127659574468086, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.0001352752086549095, | |
| "loss": 3.0506, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 3.2340425531914896, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 0.0001344707383987934, | |
| "loss": 3.0533, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 3.25531914893617, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00013366372842469105, | |
| "loss": 3.0211, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 3.276595744680851, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.0001328542381910835, | |
| "loss": 3.1129, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 3.297872340425532, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00013204232733919112, | |
| "loss": 3.0158, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.3191489361702127, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00013122805568857948, | |
| "loss": 3.0605, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.3404255319148937, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001304114832327518, | |
| "loss": 2.9792, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 3.3617021276595747, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00012959267013472892, | |
| "loss": 3.0647, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 3.382978723404255, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.0001287716767226167, | |
| "loss": 2.9722, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 3.404255319148936, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00012794856348516095, | |
| "loss": 3.0233, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.425531914893617, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.000127123391067291, | |
| "loss": 3.0216, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 3.4468085106382977, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00012629622026565147, | |
| "loss": 3.0703, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 3.4680851063829787, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00012546711202412287, | |
| "loss": 3.0121, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 3.4893617021276597, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00012463612742933148, | |
| "loss": 3.0189, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 3.5106382978723403, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00012380332770614856, | |
| "loss": 2.9589, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.5319148936170213, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.0001229687742131796, | |
| "loss": 2.9954, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 3.5531914893617023, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00012213252843824325, | |
| "loss": 3.0266, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 3.574468085106383, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00012129465199384157, | |
| "loss": 3.0273, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.595744680851064, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.0001204552066126201, | |
| "loss": 3.0214, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 3.617021276595745, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001196142541428197, | |
| "loss": 3.0232, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.6382978723404253, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00011877185654371987, | |
| "loss": 3.0004, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 3.6595744680851063, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00011792807588107357, | |
| "loss": 3.0165, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.6808510638297873, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.00011708297432253444, | |
| "loss": 3.0491, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 3.702127659574468, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00011623661413307639, | |
| "loss": 2.9456, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.723404255319149, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.0001153890576704062, | |
| "loss": 3.0586, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.74468085106383, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00011454036738036899, | |
| "loss": 3.0125, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.7659574468085104, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00011369060579234754, | |
| "loss": 2.9722, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.7872340425531914, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00011283983551465511, | |
| "loss": 2.9201, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.8085106382978724, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00011198811922992274, | |
| "loss": 3.0565, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.829787234042553, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00011113551969048089, | |
| "loss": 3.0615, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.851063829787234, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00011028209971373605, | |
| "loss": 3.0731, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.872340425531915, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00010942792217754245, | |
| "loss": 3.0144, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.8936170212765955, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00010857305001556944, | |
| "loss": 2.9905, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.9148936170212765, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00010771754621266466, | |
| "loss": 3.0232, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.9361702127659575, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00010686147380021342, | |
| "loss": 3.0408, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.9574468085106385, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00010600489585149484, | |
| "loss": 2.9963, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.978723404255319, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 0.00010514787547703466, | |
| "loss": 3.0049, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.00010429047581995546, | |
| "loss": 2.4433, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 4.0212765957446805, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00010343276005132436, | |
| "loss": 2.8295, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 4.042553191489362, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 0.00010257479136549889, | |
| "loss": 2.8904, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.0638297872340425, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 0.00010171663297547076, | |
| "loss": 2.834, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 4.085106382978723, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00010085834810820871, | |
| "loss": 2.9309, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 4.1063829787234045, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.0001, | |
| "loss": 2.9461, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 4.127659574468085, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 9.914165189179131e-05, | |
| "loss": 2.9405, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 4.148936170212766, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 9.828336702452927e-05, | |
| "loss": 2.7445, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 4.170212765957447, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 9.742520863450115e-05, | |
| "loss": 2.963, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 4.191489361702128, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 9.656723994867566e-05, | |
| "loss": 2.8778, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 4.212765957446808, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 9.570952418004455e-05, | |
| "loss": 2.9148, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 4.23404255319149, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 9.485212452296535e-05, | |
| "loss": 2.9028, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 4.25531914893617, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 9.399510414850518e-05, | |
| "loss": 2.898, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.276595744680851, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 9.313852619978659e-05, | |
| "loss": 2.9883, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 4.297872340425532, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 9.228245378733537e-05, | |
| "loss": 2.886, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 4.319148936170213, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 9.142694998443056e-05, | |
| "loss": 2.9453, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 4.340425531914893, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 9.057207782245757e-05, | |
| "loss": 2.8555, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 4.361702127659575, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 8.971790028626395e-05, | |
| "loss": 2.9359, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 4.382978723404255, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 8.886448030951912e-05, | |
| "loss": 2.8469, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 4.404255319148936, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 8.801188077007728e-05, | |
| "loss": 2.8963, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 4.425531914893617, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 8.71601644853449e-05, | |
| "loss": 2.8965, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 4.446808510638298, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 8.630939420765247e-05, | |
| "loss": 2.9457, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 4.468085106382979, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 8.545963261963102e-05, | |
| "loss": 2.8918, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.48936170212766, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 8.461094232959381e-05, | |
| "loss": 2.8957, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 4.51063829787234, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 8.376338586692366e-05, | |
| "loss": 2.8224, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 4.531914893617021, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 8.29170256774656e-05, | |
| "loss": 2.859, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 4.553191489361702, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.207192411892646e-05, | |
| "loss": 2.8896, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 4.574468085106383, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.122814345628016e-05, | |
| "loss": 2.8874, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.595744680851064, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 8.038574585718032e-05, | |
| "loss": 2.8869, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 4.617021276595745, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 7.954479338737995e-05, | |
| "loss": 2.8923, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 4.638297872340425, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 7.870534800615845e-05, | |
| "loss": 2.868, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 4.659574468085106, | |
| "grad_norm": 0.625, | |
| "learning_rate": 7.786747156175676e-05, | |
| "loss": 2.8831, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 4.680851063829787, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 7.703122578682046e-05, | |
| "loss": 2.9084, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.702127659574468, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 7.619667229385146e-05, | |
| "loss": 2.8085, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 4.723404255319149, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 7.536387257066854e-05, | |
| "loss": 2.92, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 4.74468085106383, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 7.453288797587714e-05, | |
| "loss": 2.8661, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 4.76595744680851, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 7.370377973434855e-05, | |
| "loss": 2.8322, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.787234042553192, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 7.2876608932709e-05, | |
| "loss": 2.772, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.808510638297872, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 7.205143651483906e-05, | |
| "loss": 2.905, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 4.829787234042553, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 7.122832327738331e-05, | |
| "loss": 2.9116, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 4.851063829787234, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 7.040732986527108e-05, | |
| "loss": 2.9203, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 4.872340425531915, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 6.958851676724823e-05, | |
| "loss": 2.8646, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 4.8936170212765955, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 6.877194431142055e-05, | |
| "loss": 2.844, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.914893617021277, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 6.79576726608089e-05, | |
| "loss": 2.8604, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 4.9361702127659575, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 6.714576180891654e-05, | |
| "loss": 2.8686, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 4.957446808510638, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 6.633627157530899e-05, | |
| "loss": 2.8085, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 4.9787234042553195, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 6.552926160120663e-05, | |
| "loss": 2.8017, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 6.472479134509052e-05, | |
| "loss": 2.189, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 5.0212765957446805, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 6.392292007832168e-05, | |
| "loss": 2.7068, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 5.042553191489362, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 6.312370688077399e-05, | |
| "loss": 2.7591, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 5.0638297872340425, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 6.232721063648148e-05, | |
| "loss": 2.7161, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 5.085106382978723, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 6.153349002929987e-05, | |
| "loss": 2.8126, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 5.1063829787234045, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 6.0742603538582835e-05, | |
| "loss": 2.8485, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.127659574468085, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 5.9954609434873344e-05, | |
| "loss": 2.8336, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 5.148936170212766, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 5.9169565775610656e-05, | |
| "loss": 2.6482, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 5.170212765957447, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 5.838753040085256e-05, | |
| "loss": 2.8597, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 5.191489361702128, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 5.7608560929013946e-05, | |
| "loss": 2.7875, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 5.212765957446808, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 5.683271475262164e-05, | |
| "loss": 2.822, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 5.23404255319149, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 5.6060049034085815e-05, | |
| "loss": 2.8034, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 5.25531914893617, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 5.5290620701488594e-05, | |
| "loss": 2.7899, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 5.276595744680851, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 5.452448644438946e-05, | |
| "loss": 2.8848, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 5.297872340425532, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 5.3761702709648556e-05, | |
| "loss": 2.7907, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 5.319148936170213, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 5.300232569726804e-05, | |
| "loss": 2.8616, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.340425531914893, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 5.224641135625119e-05, | |
| "loss": 2.7745, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 5.361702127659575, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 5.1494015380480396e-05, | |
| "loss": 2.8555, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 5.382978723404255, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 5.074519320461357e-05, | |
| "loss": 2.7636, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 5.404255319148936, | |
| "grad_norm": 0.625, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 2.8076, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 5.425531914893617, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 4.9258490670615475e-05, | |
| "loss": 2.8087, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 5.446808510638298, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 4.852071984901696e-05, | |
| "loss": 2.8507, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 5.468085106382979, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 4.778674189231751e-05, | |
| "loss": 2.7981, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 5.48936170212766, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 4.7056610878181486e-05, | |
| "loss": 2.8039, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 5.51063829787234, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 4.633038060083996e-05, | |
| "loss": 2.7239, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 5.531914893617021, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 4.560810456712754e-05, | |
| "loss": 2.7612, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.553191489361702, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 4.488983599254001e-05, | |
| "loss": 2.7895, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 5.574468085106383, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 4.417562779731355e-05, | |
| "loss": 2.7883, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 5.595744680851064, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 4.346553260252574e-05, | |
| "loss": 2.7913, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 5.617021276595745, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 4.275960272621852e-05, | |
| "loss": 2.7905, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 5.638297872340425, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 4.205789017954364e-05, | |
| "loss": 2.7663, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 5.659574468085106, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 4.136044666293044e-05, | |
| "loss": 2.7839, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 5.680851063829787, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 4.0667323562276814e-05, | |
| "loss": 2.7986, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 5.702127659574468, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 3.997857194516319e-05, | |
| "loss": 2.7071, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 5.723404255319149, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 3.929424255708999e-05, | |
| "loss": 2.8141, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 5.74468085106383, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 3.8614385817738794e-05, | |
| "loss": 2.7508, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.76595744680851, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 3.793905181725772e-05, | |
| "loss": 2.7273, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 5.787234042553192, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 3.726829031257062e-05, | |
| "loss": 2.6695, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 5.808510638297872, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 3.660215072371135e-05, | |
| "loss": 2.7872, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 5.829787234042553, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 3.594068213018249e-05, | |
| "loss": 2.7969, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 5.851063829787234, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 3.528393326733941e-05, | |
| "loss": 2.8035, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.872340425531915, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 3.463195252279939e-05, | |
| "loss": 2.7496, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 5.8936170212765955, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 3.3984787932876814e-05, | |
| "loss": 2.7365, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 5.914893617021277, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 3.334248717904368e-05, | |
| "loss": 2.7371, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 5.9361702127659575, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 3.270509758441671e-05, | |
| "loss": 2.7465, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 5.957446808510638, | |
| "grad_norm": 0.8828125, | |
| "learning_rate": 3.207266611027069e-05, | |
| "loss": 2.6859, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.9787234042553195, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 3.144523935257846e-05, | |
| "loss": 2.6722, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.375, | |
| "learning_rate": 3.082286353857782e-05, | |
| "loss": 2.0584, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 6.0212765957446805, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 3.0205584523365626e-05, | |
| "loss": 2.6076, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 6.042553191489362, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 2.9593447786519425e-05, | |
| "loss": 2.6513, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 6.0638297872340425, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 2.8986498428746444e-05, | |
| "loss": 2.6075, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 6.085106382978723, | |
| "grad_norm": 0.94140625, | |
| "learning_rate": 2.8384781168560693e-05, | |
| "loss": 2.7151, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 6.1063829787234045, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 2.7788340338988385e-05, | |
| "loss": 2.7812, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 6.127659574468085, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 2.719721988430153e-05, | |
| "loss": 2.7936, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 6.148936170212766, | |
| "grad_norm": 1.0, | |
| "learning_rate": 2.6611463356780096e-05, | |
| "loss": 2.6086, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 6.170212765957447, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 2.6031113913503337e-05, | |
| "loss": 2.8151, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 6.191489361702128, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 2.5456214313170002e-05, | |
| "loss": 2.7246, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 6.212765957446808, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 2.4886806912948035e-05, | |
| "loss": 2.7524, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 6.23404255319149, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 2.4322933665353776e-05, | |
| "loss": 2.7285, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 6.25531914893617, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 2.3764636115160978e-05, | |
| "loss": 2.7237, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 6.276595744680851, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 2.3211955396340002e-05, | |
| "loss": 2.818, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 6.297872340425532, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 2.2664932229027024e-05, | |
| "loss": 2.7163, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 6.319148936170213, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 2.2123606916523953e-05, | |
| "loss": 2.7859, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 6.340425531914893, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 2.1588019342328968e-05, | |
| "loss": 2.6892, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 6.361702127659575, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 2.1058208967198045e-05, | |
| "loss": 2.767, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 6.382978723404255, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 2.0534214826237484e-05, | |
| "loss": 2.6933, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.404255319148936, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 2.0016075526028065e-05, | |
| "loss": 2.7303, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 6.425531914893617, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 1.9503829241780412e-05, | |
| "loss": 2.7377, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 6.446808510638298, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 1.8997513714522487e-05, | |
| "loss": 2.7818, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 6.468085106382979, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 1.8497166248318876e-05, | |
| "loss": 2.7335, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 6.48936170212766, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 1.8002823707522297e-05, | |
| "loss": 2.733, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 6.51063829787234, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 1.7514522514057553e-05, | |
| "loss": 2.6446, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 6.531914893617021, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 1.703229864473811e-05, | |
| "loss": 2.6907, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 6.553191489361702, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 1.6556187628615273e-05, | |
| "loss": 2.7176, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 6.574468085106383, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 1.608622454436062e-05, | |
| "loss": 2.7109, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 6.595744680851064, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 1.562244401768144e-05, | |
| "loss": 2.7085, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 6.617021276595745, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 1.5164880218769618e-05, | |
| "loss": 2.6987, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 6.638297872340425, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 1.4713566859784045e-05, | |
| "loss": 2.6835, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 6.659574468085106, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 1.426853719236676e-05, | |
| "loss": 2.6981, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 6.680851063829787, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 1.3829824005193181e-05, | |
| "loss": 2.7132, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 6.702127659574468, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 1.339745962155613e-05, | |
| "loss": 2.6319, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 6.723404255319149, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 1.2971475896984475e-05, | |
| "loss": 2.7332, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 6.74468085106383, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 1.2551904216896037e-05, | |
| "loss": 2.6649, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 6.76595744680851, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 1.2138775494285182e-05, | |
| "loss": 2.6486, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 6.787234042553192, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 1.1732120167445248e-05, | |
| "loss": 2.5875, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 6.808510638297872, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 1.1331968197725984e-05, | |
| "loss": 2.7079, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.829787234042553, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 1.0938349067325959e-05, | |
| "loss": 2.7134, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 6.851063829787234, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 1.0551291777120464e-05, | |
| "loss": 2.7199, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 6.872340425531915, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 1.0170824844524728e-05, | |
| "loss": 2.6655, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 6.8936170212765955, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 9.796976301392934e-06, | |
| "loss": 2.6519, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 6.914893617021277, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 9.429773691952858e-06, | |
| "loss": 2.6443, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.9361702127659575, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 9.069244070776428e-06, | |
| "loss": 2.6531, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 6.957446808510638, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 8.715414000786448e-06, | |
| "loss": 2.5897, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 6.9787234042553195, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 8.368309551299536e-06, | |
| "loss": 2.5772, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 8.027956296105354e-06, | |
| "loss": 1.9731, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 7.0212765957446805, | |
| "grad_norm": 0.75, | |
| "learning_rate": 7.6943793115824e-06, | |
| "loss": 2.5669, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 7.042553191489362, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 7.367603174850502e-06, | |
| "loss": 2.6154, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 7.0638297872340425, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 7.047651961959978e-06, | |
| "loss": 2.5542, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 7.085106382978723, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 6.73454924611776e-06, | |
| "loss": 2.6428, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 7.1063829787234045, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 6.428318095950647e-06, | |
| "loss": 2.6929, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 7.127659574468085, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 6.128981073805584e-06, | |
| "loss": 2.6994, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 7.148936170212766, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 5.836560234087418e-06, | |
| "loss": 2.5162, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 7.170212765957447, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 5.551077121633874e-06, | |
| "loss": 2.7308, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 7.191489361702128, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 5.272552770128314e-06, | |
| "loss": 2.6655, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 7.212765957446808, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 5.001007700549898e-06, | |
| "loss": 2.7014, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 7.23404255319149, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 4.7364619196617495e-06, | |
| "loss": 2.6704, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 7.25531914893617, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 4.478934918536837e-06, | |
| "loss": 2.6756, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 7.276595744680851, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 4.228445671121972e-06, | |
| "loss": 2.7574, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 7.297872340425532, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 3.985012632839824e-06, | |
| "loss": 2.6565, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 7.319148936170213, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 3.748653739229191e-06, | |
| "loss": 2.7389, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 7.340425531914893, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 3.519386404623537e-06, | |
| "loss": 2.6382, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 7.361702127659575, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 3.2972275208679625e-06, | |
| "loss": 2.7147, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 7.382978723404255, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 3.0821934560746447e-06, | |
| "loss": 2.6497, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 7.404255319148936, | |
| "grad_norm": 0.76171875, | |
| "learning_rate": 2.8743000534168675e-06, | |
| "loss": 2.6844, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 7.425531914893617, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 2.6735626299617457e-06, | |
| "loss": 2.6961, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 7.446808510638298, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 2.479995975541749e-06, | |
| "loss": 2.7341, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.468085106382979, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 2.2936143516649188e-06, | |
| "loss": 2.6872, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 7.48936170212766, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 2.1144314904642195e-06, | |
| "loss": 2.6879, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 7.51063829787234, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 1.942460593685713e-06, | |
| "loss": 2.5916, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 7.531914893617021, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 1.7777143317159406e-06, | |
| "loss": 2.643, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 7.553191489361702, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 1.6202048426483651e-06, | |
| "loss": 2.6724, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 7.574468085106383, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 1.4699437313891007e-06, | |
| "loss": 2.6634, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 7.595744680851064, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 1.3269420688018508e-06, | |
| "loss": 2.6651, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 7.617021276595745, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 1.1912103908922945e-06, | |
| "loss": 2.6545, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 7.638297872340425, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 1.0627586980317073e-06, | |
| "loss": 2.6455, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 7.659574468085106, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 9.415964542203059e-07, | |
| "loss": 2.6622, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 7.680851063829787, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 8.277325863898511e-07, | |
| "loss": 2.6787, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 7.702127659574468, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 7.21175483745995e-07, | |
| "loss": 2.6004, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 7.723404255319149, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 6.219329971501653e-07, | |
| "loss": 2.7023, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 7.74468085106383, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 5.300124385410943e-07, | |
| "loss": 2.6309, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 7.76595744680851, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 4.4542058039619417e-07, | |
| "loss": 2.6197, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 7.787234042553192, | |
| "grad_norm": 0.75, | |
| "learning_rate": 3.681636552324452e-07, | |
| "loss": 2.5579, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 7.808510638297872, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 2.9824735514732974e-07, | |
| "loss": 2.6765, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 7.829787234042553, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 2.3567683139936735e-07, | |
| "loss": 2.687, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 7.851063829787234, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 1.8045669402859677e-07, | |
| "loss": 2.6924, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 7.872340425531915, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 1.3259101151694708e-07, | |
| "loss": 2.6409, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 7.8936170212765955, | |
| "grad_norm": 0.78125, | |
| "learning_rate": 9.208331048846663e-08, | |
| "loss": 2.6251, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 7.914893617021277, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 5.893657544947528e-08, | |
| "loss": 2.616, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 7.9361702127659575, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 3.3153248568695835e-08, | |
| "loss": 2.626, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 7.957446808510638, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 1.47352294973091e-08, | |
| "loss": 2.5585, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 7.9787234042553195, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 3.6838752290102585e-09, | |
| "loss": 2.543, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.0, | |
| "loss": 1.9425, | |
| "step": 376 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 376, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 47, | |
| "total_flos": 5.59507839123456e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |