pszemraj's picture
add additional 2-epoch checkpoint, better regularization
d4ed403
raw
history blame
7.06 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.991123701605288,
"global_step": 164,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04,
"learning_rate": 0.00030000000000000003,
"loss": 2.5934,
"step": 3
},
{
"epoch": 0.07,
"learning_rate": 0.0003998458072481446,
"loss": 2.5887,
"step": 6
},
{
"epoch": 0.11,
"learning_rate": 0.0003990369453344394,
"loss": 2.5383,
"step": 9
},
{
"epoch": 0.15,
"learning_rate": 0.00039753766811902755,
"loss": 2.5382,
"step": 12
},
{
"epoch": 0.18,
"learning_rate": 0.0003953531762641745,
"loss": 2.5684,
"step": 15
},
{
"epoch": 0.22,
"learning_rate": 0.00039249104729072946,
"loss": 2.5694,
"step": 18
},
{
"epoch": 0.25,
"learning_rate": 0.00038896120929337566,
"loss": 2.5685,
"step": 21
},
{
"epoch": 0.29,
"learning_rate": 0.0003847759065022574,
"loss": 2.5249,
"step": 24
},
{
"epoch": 0.33,
"learning_rate": 0.00037994965681044433,
"loss": 2.5835,
"step": 27
},
{
"epoch": 0.36,
"learning_rate": 0.00037449920141455944,
"loss": 2.5326,
"step": 30
},
{
"epoch": 0.4,
"learning_rate": 0.00036844344674325733,
"loss": 2.564,
"step": 33
},
{
"epoch": 0.44,
"learning_rate": 0.0003618033988749895,
"loss": 2.5717,
"step": 36
},
{
"epoch": 0.47,
"learning_rate": 0.0003546020906725474,
"loss": 2.5149,
"step": 39
},
{
"epoch": 0.51,
"learning_rate": 0.0003468645018871371,
"loss": 2.5695,
"step": 42
},
{
"epoch": 0.54,
"learning_rate": 0.0003386174725091272,
"loss": 2.5374,
"step": 45
},
{
"epoch": 0.58,
"learning_rate": 0.0003298896096660367,
"loss": 2.5413,
"step": 48
},
{
"epoch": 0.62,
"learning_rate": 0.0003207111883907143,
"loss": 2.509,
"step": 51
},
{
"epoch": 0.65,
"learning_rate": 0.00031111404660392046,
"loss": 2.5628,
"step": 54
},
{
"epoch": 0.69,
"learning_rate": 0.00030113147467559695,
"loss": 2.5069,
"step": 57
},
{
"epoch": 0.73,
"learning_rate": 0.00029079809994790937,
"loss": 2.5331,
"step": 60
},
{
"epoch": 0.76,
"learning_rate": 0.0002801497666206282,
"loss": 2.5527,
"step": 63
},
{
"epoch": 0.8,
"learning_rate": 0.0002692234114154986,
"loss": 2.6179,
"step": 66
},
{
"epoch": 0.83,
"learning_rate": 0.00025805693545089247,
"loss": 2.5411,
"step": 69
},
{
"epoch": 0.87,
"learning_rate": 0.00024668907277118114,
"loss": 2.5583,
"step": 72
},
{
"epoch": 0.91,
"learning_rate": 0.00023515925598687094,
"loss": 2.534,
"step": 75
},
{
"epoch": 0.94,
"learning_rate": 0.00022350747949156756,
"loss": 2.5433,
"step": 78
},
{
"epoch": 0.98,
"learning_rate": 0.0002117741607302378,
"loss": 2.5487,
"step": 81
},
{
"epoch": 1.02,
"learning_rate": 0.0002,
"loss": 3.1324,
"step": 84
},
{
"epoch": 1.06,
"learning_rate": 0.00018822583926976218,
"loss": 2.4641,
"step": 87
},
{
"epoch": 1.1,
"learning_rate": 0.00017649252050843252,
"loss": 2.4806,
"step": 90
},
{
"epoch": 1.13,
"learning_rate": 0.0001648407440131291,
"loss": 2.4693,
"step": 93
},
{
"epoch": 1.17,
"learning_rate": 0.000153310927228819,
"loss": 2.5011,
"step": 96
},
{
"epoch": 1.21,
"learning_rate": 0.00014194306454910757,
"loss": 2.4595,
"step": 99
},
{
"epoch": 1.24,
"learning_rate": 0.00013077658858450138,
"loss": 2.4893,
"step": 102
},
{
"epoch": 1.28,
"learning_rate": 0.00011985023337937184,
"loss": 2.4932,
"step": 105
},
{
"epoch": 1.31,
"learning_rate": 0.00010920190005209065,
"loss": 2.4871,
"step": 108
},
{
"epoch": 1.35,
"learning_rate": 9.886852532440312e-05,
"loss": 2.4672,
"step": 111
},
{
"epoch": 1.39,
"learning_rate": 8.888595339607961e-05,
"loss": 2.4597,
"step": 114
},
{
"epoch": 1.42,
"learning_rate": 7.928881160928572e-05,
"loss": 2.4531,
"step": 117
},
{
"epoch": 1.46,
"learning_rate": 7.011039033396329e-05,
"loss": 2.4749,
"step": 120
},
{
"epoch": 1.5,
"learning_rate": 6.138252749087286e-05,
"loss": 2.4551,
"step": 123
},
{
"epoch": 1.53,
"learning_rate": 5.313549811286293e-05,
"loss": 2.4796,
"step": 126
},
{
"epoch": 1.57,
"learning_rate": 4.53979093274526e-05,
"loss": 2.4705,
"step": 129
},
{
"epoch": 1.6,
"learning_rate": 3.819660112501053e-05,
"loss": 2.4915,
"step": 132
},
{
"epoch": 1.64,
"learning_rate": 3.1556553256742713e-05,
"loss": 2.4958,
"step": 135
},
{
"epoch": 1.68,
"learning_rate": 2.5500798585440567e-05,
"loss": 2.4814,
"step": 138
},
{
"epoch": 1.71,
"learning_rate": 2.0050343189555743e-05,
"loss": 2.5034,
"step": 141
},
{
"epoch": 1.75,
"learning_rate": 1.5224093497742653e-05,
"loss": 2.4671,
"step": 144
},
{
"epoch": 1.79,
"learning_rate": 1.1038790706624391e-05,
"loss": 2.4987,
"step": 147
},
{
"epoch": 1.82,
"learning_rate": 7.508952709270567e-06,
"loss": 2.4618,
"step": 150
},
{
"epoch": 1.86,
"learning_rate": 4.646823735825523e-06,
"loss": 2.4815,
"step": 153
},
{
"epoch": 1.89,
"learning_rate": 2.462331880972468e-06,
"loss": 2.4595,
"step": 156
},
{
"epoch": 1.93,
"learning_rate": 9.630546655606364e-07,
"loss": 2.4925,
"step": 159
},
{
"epoch": 1.97,
"learning_rate": 1.5419275185541982e-07,
"loss": 2.4333,
"step": 162
},
{
"epoch": 1.99,
"step": 164,
"total_flos": 4.620604962546647e+17,
"train_loss": 2.525294606278582,
"train_runtime": 28715.0205,
"train_samples_per_second": 0.738,
"train_steps_per_second": 0.006
}
],
"max_steps": 164,
"num_train_epochs": 2,
"total_flos": 4.620604962546647e+17,
"trial_name": null,
"trial_params": null
}