| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 188, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02127659574468085, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 2e-05, | |
| "loss": 3.872, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0425531914893617, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 4e-05, | |
| "loss": 3.9714, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06382978723404255, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 6e-05, | |
| "loss": 3.9502, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0851063829787234, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 8e-05, | |
| "loss": 4.0793, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.10638297872340426, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 0.0001, | |
| "loss": 3.9545, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1276595744680851, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00012, | |
| "loss": 3.803, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14893617021276595, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.00014, | |
| "loss": 3.5993, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1702127659574468, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 0.00016, | |
| "loss": 3.8622, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.19148936170212766, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 0.00018, | |
| "loss": 3.6551, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2127659574468085, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.0002, | |
| "loss": 3.6643, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.23404255319148937, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00019998442534369985, | |
| "loss": 3.7215, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2553191489361702, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00019993770622619782, | |
| "loss": 3.6194, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2765957446808511, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00019985985720017785, | |
| "loss": 3.7064, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2978723404255319, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 0.00019975090251507638, | |
| "loss": 3.5715, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.3191489361702128, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.0001996108761095289, | |
| "loss": 3.5818, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3404255319148936, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001994398216007982, | |
| "loss": 3.5229, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3617021276595745, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.0001992377922711879, | |
| "loss": 3.5894, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3829787234042553, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00019900485105144543, | |
| "loss": 3.5212, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.40425531914893614, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 0.00019874107050115954, | |
| "loss": 3.5943, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.425531914893617, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 0.00019844653278615833, | |
| "loss": 3.5288, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.44680851063829785, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 0.00019812132965291545, | |
| "loss": 3.6237, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.46808510638297873, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 0.00019776556239997146, | |
| "loss": 3.5176, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.48936170212765956, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00019737934184638006, | |
| "loss": 3.5751, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.5106382978723404, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 0.00019696278829718883, | |
| "loss": 3.5265, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5319148936170213, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00019651603150596495, | |
| "loss": 3.5635, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5531914893617021, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 0.00019603921063437793, | |
| "loss": 3.543, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.574468085106383, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 0.00019553247420885157, | |
| "loss": 3.573, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5957446808510638, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.0001949959800742991, | |
| "loss": 3.566, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.6170212765957447, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 0.00019442989534495557, | |
| "loss": 3.6022, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6382978723404256, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00019383439635232294, | |
| "loss": 3.514, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6595744680851063, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 0.00019320966859024397, | |
| "loss": 3.5634, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6808510638297872, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00019255590665712214, | |
| "loss": 3.6299, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.7021276595744681, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0001918733141953056, | |
| "loss": 3.4891, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.723404255319149, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.0001911621038276542, | |
| "loss": 3.6105, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7446808510638298, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001904224970913085, | |
| "loss": 3.5861, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7659574468085106, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00018965472436868286, | |
| "loss": 3.5116, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7872340425531915, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001888590248157027, | |
| "loss": 3.4935, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.8085106382978723, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 0.00018803564628730915, | |
| "loss": 3.6211, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.8297872340425532, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.00018718484526025387, | |
| "loss": 3.634, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.851063829787234, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00018630688675320842, | |
| "loss": 3.6452, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8723404255319149, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.00018540204424421263, | |
| "loss": 3.5727, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8936170212765957, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.0001844705995854882, | |
| "loss": 3.5772, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.9148936170212766, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00018351284291564358, | |
| "loss": 3.6298, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.9361702127659575, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00018252907256929775, | |
| "loss": 3.6612, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9574468085106383, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00018151959498415122, | |
| "loss": 3.6252, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9787234042553191, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00018048472460553257, | |
| "loss": 3.6423, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.0001794247837884511, | |
| "loss": 3.4633, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.0212765957446808, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 0.00017834010269718526, | |
| "loss": 3.2725, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.0425531914893618, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.0001772310192024389, | |
| "loss": 3.3142, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.0638297872340425, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 0.0001760978787760968, | |
| "loss": 3.3171, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0851063829787233, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.0001749410343836125, | |
| "loss": 3.4249, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.1063829787234043, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00017376084637406222, | |
| "loss": 3.37, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.127659574468085, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00017255768236789826, | |
| "loss": 3.3532, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.148936170212766, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.00017133191714243805, | |
| "loss": 3.1741, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.1702127659574468, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00017008393251512332, | |
| "loss": 3.402, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.1914893617021276, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00016881411722458688, | |
| "loss": 3.2908, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.2127659574468086, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00016752286680956306, | |
| "loss": 3.3467, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.2340425531914894, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00016621058348568007, | |
| "loss": 3.3667, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.2553191489361701, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00016487767602017263, | |
| "loss": 3.308, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.2765957446808511, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00016352455960455387, | |
| "loss": 3.4156, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.297872340425532, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00016215165572528597, | |
| "loss": 3.3147, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.3191489361702127, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 0.0001607593920324899, | |
| "loss": 3.3296, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.3404255319148937, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00015934820220673564, | |
| "loss": 3.2744, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.3617021276595744, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00015791852582395334, | |
| "loss": 3.3446, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.3829787234042552, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001564708082185087, | |
| "loss": 3.2568, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.4042553191489362, | |
| "grad_norm": 0.220703125, | |
| "learning_rate": 0.00015500550034448413, | |
| "loss": 3.319, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.425531914893617, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00015352305863520991, | |
| "loss": 3.2982, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.4468085106382977, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001520239448610882, | |
| "loss": 3.3751, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.4680851063829787, | |
| "grad_norm": 0.25, | |
| "learning_rate": 0.00015050862598575476, | |
| "loss": 3.293, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.4893617021276595, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 0.00014897757402062284, | |
| "loss": 3.3238, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.5106382978723403, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00014743126587785522, | |
| "loss": 3.2718, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.5319148936170213, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00014587018322180905, | |
| "loss": 3.3136, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.5531914893617023, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 0.00014429481231900083, | |
| "loss": 3.3282, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.574468085106383, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.0001427056438866376, | |
| "loss": 3.3317, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.5957446808510638, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00014110317293976218, | |
| "loss": 3.3342, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.6170212765957448, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00013948789863705912, | |
| "loss": 3.351, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.6382978723404256, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 0.00013786032412537035, | |
| "loss": 3.3016, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.6595744680851063, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 0.00013622095638296826, | |
| "loss": 3.3357, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.6808510638297873, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00013457030606163562, | |
| "loss": 3.3777, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.702127659574468, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.000132908887327601, | |
| "loss": 3.2683, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.7234042553191489, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.00013123721770137944, | |
| "loss": 3.3866, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.7446808510638299, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 0.00012955581789656843, | |
| "loss": 3.3582, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.7659574468085106, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001278652116576492, | |
| "loss": 3.2956, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.7872340425531914, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.0001261659255968441, | |
| "loss": 3.258, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.8085106382978724, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 0.00012445848903008003, | |
| "loss": 3.4091, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.8297872340425532, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 0.00012274343381211066, | |
| "loss": 3.4071, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.851063829787234, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 0.00012102129417084714, | |
| "loss": 3.4251, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.872340425531915, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00011929260654094969, | |
| "loss": 3.3607, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.8936170212765957, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00011755790939673209, | |
| "loss": 3.3509, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.9148936170212765, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001158177430844304, | |
| "loss": 3.4081, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.9361702127659575, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00011407264965388906, | |
| "loss": 3.4399, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.9574468085106385, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00011232317268971585, | |
| "loss": 3.4059, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.978723404255319, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00011056985714195932, | |
| "loss": 3.4202, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.84375, | |
| "learning_rate": 0.00010881324915636019, | |
| "loss": 3.0814, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.021276595744681, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0001070538959042311, | |
| "loss": 3.1079, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0425531914893615, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00010529234541201631, | |
| "loss": 3.1756, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0638297872340425, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00010352914639058526, | |
| "loss": 3.1467, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.0851063829787235, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00010176484806431288, | |
| "loss": 3.2387, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.106382978723404, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.0001, | |
| "loss": 3.2315, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.127659574468085, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.823515193568715e-05, | |
| "loss": 3.2065, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.148936170212766, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.647085360941476e-05, | |
| "loss": 3.0125, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.1702127659574466, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.470765458798368e-05, | |
| "loss": 3.25, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.1914893617021276, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 9.29461040957689e-05, | |
| "loss": 3.1501, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.2127659574468086, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 9.118675084363986e-05, | |
| "loss": 3.1963, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.2340425531914896, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 8.943014285804072e-05, | |
| "loss": 3.205, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.25531914893617, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.767682731028415e-05, | |
| "loss": 3.1699, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.276595744680851, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 8.592735034611097e-05, | |
| "loss": 3.2659, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.297872340425532, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 8.418225691556962e-05, | |
| "loss": 3.1723, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.3191489361702127, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 8.244209060326794e-05, | |
| "loss": 3.1985, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.3404255319148937, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 8.070739345905032e-05, | |
| "loss": 3.1365, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.3617021276595747, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 7.897870582915288e-05, | |
| "loss": 3.2121, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.382978723404255, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 7.725656618788937e-05, | |
| "loss": 3.115, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.404255319148936, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 7.554151096992001e-05, | |
| "loss": 3.1752, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.425531914893617, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.383407440315596e-05, | |
| "loss": 3.1718, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.4468085106382977, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.213478834235079e-05, | |
| "loss": 3.2311, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.4680851063829787, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 7.04441821034316e-05, | |
| "loss": 3.167, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.4893617021276597, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 6.87627822986206e-05, | |
| "loss": 3.1866, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.5106382978723403, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 6.7091112672399e-05, | |
| "loss": 3.1326, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.5319148936170213, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 6.542969393836436e-05, | |
| "loss": 3.1708, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.5531914893617023, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 6.377904361703178e-05, | |
| "loss": 3.1963, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.574468085106383, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 6.213967587462968e-05, | |
| "loss": 3.1986, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.595744680851064, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 6.051210136294089e-05, | |
| "loss": 3.1945, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.617021276595745, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 5.889682706023783e-05, | |
| "loss": 3.2107, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.6382978723404253, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 5.729435611336239e-05, | |
| "loss": 3.1784, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.6595744680851063, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 5.570518768099918e-05, | |
| "loss": 3.1984, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6808510638297873, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 5.4129816778190936e-05, | |
| "loss": 3.2371, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.702127659574468, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 5.2568734122144756e-05, | |
| "loss": 3.1334, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.723404255319149, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 5.102242597937717e-05, | |
| "loss": 3.255, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.74468085106383, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 4.949137401424527e-05, | |
| "loss": 3.2194, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.7659574468085104, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 4.797605513891179e-05, | |
| "loss": 3.1648, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.7872340425531914, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 4.6476941364790074e-05, | |
| "loss": 3.1241, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.8085106382978724, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 4.4994499655515865e-05, | |
| "loss": 3.2732, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.829787234042553, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 4.352919178149132e-05, | |
| "loss": 3.2714, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.851063829787234, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 4.2081474176046646e-05, | |
| "loss": 3.2881, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.872340425531915, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 4.0651797793264354e-05, | |
| "loss": 3.2259, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.8936170212765955, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.924060796751012e-05, | |
| "loss": 3.2168, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.9148936170212765, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 3.784834427471408e-05, | |
| "loss": 3.2721, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.9361702127659575, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 3.647544039544615e-05, | |
| "loss": 3.2962, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.9574468085106385, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 3.5122323979827395e-05, | |
| "loss": 3.2635, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.978723404255319, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.378941651431996e-05, | |
| "loss": 3.2729, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 3.2477133190436945e-05, | |
| "loss": 2.838, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 3.021276595744681, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 3.118588277541312e-05, | |
| "loss": 3.0105, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 3.0425531914893615, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 2.99160674848767e-05, | |
| "loss": 3.0744, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 3.0638297872340425, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 2.8668082857562005e-05, | |
| "loss": 3.0395, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.0851063829787235, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.7442317632101745e-05, | |
| "loss": 3.1344, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.106382978723404, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.6239153625937784e-05, | |
| "loss": 3.1349, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 3.127659574468085, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 2.5058965616387498e-05, | |
| "loss": 3.1193, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 3.148936170212766, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 2.390212122390323e-05, | |
| "loss": 2.9173, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 3.1702127659574466, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 2.2768980797561124e-05, | |
| "loss": 3.1502, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 3.1914893617021276, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 2.1659897302814747e-05, | |
| "loss": 3.065, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.2127659574468086, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 2.0575216211548907e-05, | |
| "loss": 3.1079, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 3.2340425531914896, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 1.9515275394467446e-05, | |
| "loss": 3.1094, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 3.25531914893617, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 1.8480405015848824e-05, | |
| "loss": 3.0872, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 3.276595744680851, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 1.7470927430702277e-05, | |
| "loss": 3.1752, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 3.297872340425532, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 1.648715708435645e-05, | |
| "loss": 3.0836, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.3191489361702127, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 1.5529400414511806e-05, | |
| "loss": 3.1217, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.3404255319148937, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 1.4597955755787373e-05, | |
| "loss": 3.0535, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 3.3617021276595747, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 1.3693113246791589e-05, | |
| "loss": 3.1319, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 3.382978723404255, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 1.2815154739746138e-05, | |
| "loss": 3.0385, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 3.404255319148936, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 1.196435371269089e-05, | |
| "loss": 3.0892, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.425531914893617, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 1.1140975184297331e-05, | |
| "loss": 3.1002, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 3.4468085106382977, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 1.0345275631317163e-05, | |
| "loss": 3.1541, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 3.4680851063829787, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 9.577502908691526e-06, | |
| "loss": 3.0974, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 3.4893617021276597, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 8.837896172345827e-06, | |
| "loss": 3.1121, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 3.5106382978723403, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 8.1266858046944e-06, | |
| "loss": 3.0568, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.5319148936170213, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 7.4440933428779e-06, | |
| "loss": 3.0998, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 3.5531914893617023, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 6.7903314097560454e-06, | |
| "loss": 3.1263, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 3.574468085106383, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 6.165603647677054e-06, | |
| "loss": 3.1272, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.595744680851064, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 5.570104655044428e-06, | |
| "loss": 3.1268, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 3.617021276595745, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 5.00401992570092e-06, | |
| "loss": 3.1398, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.6382978723404253, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 4.467525791148453e-06, | |
| "loss": 3.1151, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 3.6595744680851063, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 3.960789365622075e-06, | |
| "loss": 3.1346, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.6808510638297873, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 3.483968494035039e-06, | |
| "loss": 3.1692, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 3.702127659574468, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 3.0372117028111825e-06, | |
| "loss": 3.0739, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.723404255319149, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 2.6206581536199594e-06, | |
| "loss": 3.1945, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.74468085106383, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 2.2344376000285604e-06, | |
| "loss": 3.1558, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.7659574468085104, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 1.8786703470845547e-06, | |
| "loss": 3.1134, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.7872340425531914, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 1.553467213841664e-06, | |
| "loss": 3.0711, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.8085106382978724, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.2589294988404888e-06, | |
| "loss": 3.2178, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.829787234042553, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.951489485545695e-07, | |
| "loss": 3.2206, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.851063829787234, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 7.622077288121033e-07, | |
| "loss": 3.24, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.872340425531915, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 5.60178399201805e-07, | |
| "loss": 3.1806, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.8936170212765955, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.8912389047108813e-07, | |
| "loss": 3.1704, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.9148936170212765, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 2.490974849236216e-07, | |
| "loss": 3.2246, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.9361702127659575, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 1.401427998221627e-07, | |
| "loss": 3.2522, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.9574468085106385, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 6.229377380218005e-08, | |
| "loss": 3.2196, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.978723404255319, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 1.5574656300143542e-08, | |
| "loss": 3.2318, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.0, | |
| "loss": 2.7607, | |
| "step": 188 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 188, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 24, | |
| "total_flos": 2.79753919561728e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |