464 lines
12 KiB
JSON
464 lines
12 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 5.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 60,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.0903954802259887,
|
||
|
|
"grad_norm": 72.95352172851562,
|
||
|
|
"learning_rate": 0.0,
|
||
|
|
"loss": 3.539058208465576,
|
||
|
|
"step": 1
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1807909604519774,
|
||
|
|
"grad_norm": 84.62841033935547,
|
||
|
|
"learning_rate": 1.6666666666666667e-06,
|
||
|
|
"loss": 3.510171890258789,
|
||
|
|
"step": 2
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2711864406779661,
|
||
|
|
"grad_norm": 72.88843536376953,
|
||
|
|
"learning_rate": 3.3333333333333333e-06,
|
||
|
|
"loss": 3.4764676094055176,
|
||
|
|
"step": 3
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3615819209039548,
|
||
|
|
"grad_norm": 68.25614929199219,
|
||
|
|
"learning_rate": 5e-06,
|
||
|
|
"loss": 3.0605616569519043,
|
||
|
|
"step": 4
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4519774011299435,
|
||
|
|
"grad_norm": 28.142393112182617,
|
||
|
|
"learning_rate": 6.666666666666667e-06,
|
||
|
|
"loss": 2.3392772674560547,
|
||
|
|
"step": 5
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5423728813559322,
|
||
|
|
"grad_norm": 8.023438453674316,
|
||
|
|
"learning_rate": 8.333333333333334e-06,
|
||
|
|
"loss": 1.9946647882461548,
|
||
|
|
"step": 6
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.632768361581921,
|
||
|
|
"grad_norm": 6.666532039642334,
|
||
|
|
"learning_rate": 1e-05,
|
||
|
|
"loss": 1.864429235458374,
|
||
|
|
"step": 7
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7231638418079096,
|
||
|
|
"grad_norm": 5.340750694274902,
|
||
|
|
"learning_rate": 9.991540791356342e-06,
|
||
|
|
"loss": 1.7559094429016113,
|
||
|
|
"step": 8
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8135593220338984,
|
||
|
|
"grad_norm": 4.030279159545898,
|
||
|
|
"learning_rate": 9.966191788709716e-06,
|
||
|
|
"loss": 1.7070744037628174,
|
||
|
|
"step": 9
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.903954802259887,
|
||
|
|
"grad_norm": 2.765150308609009,
|
||
|
|
"learning_rate": 9.924038765061042e-06,
|
||
|
|
"loss": 1.576987624168396,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9943502824858758,
|
||
|
|
"grad_norm": 2.4855945110321045,
|
||
|
|
"learning_rate": 9.86522435289912e-06,
|
||
|
|
"loss": 1.5758938789367676,
|
||
|
|
"step": 11
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"grad_norm": 5.696783065795898,
|
||
|
|
"learning_rate": 9.789947561577445e-06,
|
||
|
|
"loss": 1.3263113498687744,
|
||
|
|
"step": 12
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0903954802259888,
|
||
|
|
"grad_norm": 2.779576539993286,
|
||
|
|
"learning_rate": 9.698463103929542e-06,
|
||
|
|
"loss": 1.5225284099578857,
|
||
|
|
"step": 13
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1807909604519775,
|
||
|
|
"grad_norm": 2.253300666809082,
|
||
|
|
"learning_rate": 9.591080534401371e-06,
|
||
|
|
"loss": 1.488105297088623,
|
||
|
|
"step": 14
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.271186440677966,
|
||
|
|
"grad_norm": 1.884597897529602,
|
||
|
|
"learning_rate": 9.468163201617063e-06,
|
||
|
|
"loss": 1.4530308246612549,
|
||
|
|
"step": 15
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3615819209039548,
|
||
|
|
"grad_norm": 1.755130648612976,
|
||
|
|
"learning_rate": 9.330127018922195e-06,
|
||
|
|
"loss": 1.422749638557434,
|
||
|
|
"step": 16
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4519774011299436,
|
||
|
|
"grad_norm": 1.560288906097412,
|
||
|
|
"learning_rate": 9.177439057064684e-06,
|
||
|
|
"loss": 1.433840036392212,
|
||
|
|
"step": 17
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.542372881355932,
|
||
|
|
"grad_norm": 1.7185840606689453,
|
||
|
|
"learning_rate": 9.01061596377522e-06,
|
||
|
|
"loss": 1.4170303344726562,
|
||
|
|
"step": 18
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6327683615819208,
|
||
|
|
"grad_norm": 1.4684430360794067,
|
||
|
|
"learning_rate": 8.83022221559489e-06,
|
||
|
|
"loss": 1.388798713684082,
|
||
|
|
"step": 19
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7231638418079096,
|
||
|
|
"grad_norm": 1.6119791269302368,
|
||
|
|
"learning_rate": 8.636868207865244e-06,
|
||
|
|
"loss": 1.3863484859466553,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8135593220338984,
|
||
|
|
"grad_norm": 1.2974920272827148,
|
||
|
|
"learning_rate": 8.43120818934367e-06,
|
||
|
|
"loss": 1.3138604164123535,
|
||
|
|
"step": 21
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9039548022598871,
|
||
|
|
"grad_norm": 1.5221943855285645,
|
||
|
|
"learning_rate": 8.213938048432697e-06,
|
||
|
|
"loss": 1.3569386005401611,
|
||
|
|
"step": 22
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9943502824858759,
|
||
|
|
"grad_norm": 1.5059224367141724,
|
||
|
|
"learning_rate": 7.985792958513932e-06,
|
||
|
|
"loss": 1.328192949295044,
|
||
|
|
"step": 23
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"grad_norm": 3.393784761428833,
|
||
|
|
"learning_rate": 7.747544890354031e-06,
|
||
|
|
"loss": 1.2474285364151,
|
||
|
|
"step": 24
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0903954802259888,
|
||
|
|
"grad_norm": 1.4098623991012573,
|
||
|
|
"learning_rate": 7.500000000000001e-06,
|
||
|
|
"loss": 1.2317988872528076,
|
||
|
|
"step": 25
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.1807909604519775,
|
||
|
|
"grad_norm": 1.3200597763061523,
|
||
|
|
"learning_rate": 7.243995901002312e-06,
|
||
|
|
"loss": 1.2224856615066528,
|
||
|
|
"step": 26
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.2711864406779663,
|
||
|
|
"grad_norm": 1.4033938646316528,
|
||
|
|
"learning_rate": 6.980398830195785e-06,
|
||
|
|
"loss": 1.2676887512207031,
|
||
|
|
"step": 27
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.361581920903955,
|
||
|
|
"grad_norm": 1.3782753944396973,
|
||
|
|
"learning_rate": 6.710100716628345e-06,
|
||
|
|
"loss": 1.2362030744552612,
|
||
|
|
"step": 28
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4519774011299433,
|
||
|
|
"grad_norm": 1.2505719661712646,
|
||
|
|
"learning_rate": 6.434016163555452e-06,
|
||
|
|
"loss": 1.2472147941589355,
|
||
|
|
"step": 29
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.542372881355932,
|
||
|
|
"grad_norm": 1.2195472717285156,
|
||
|
|
"learning_rate": 6.153079353712201e-06,
|
||
|
|
"loss": 1.198156714439392,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.632768361581921,
|
||
|
|
"grad_norm": 1.2974107265472412,
|
||
|
|
"learning_rate": 5.8682408883346535e-06,
|
||
|
|
"loss": 1.206047534942627,
|
||
|
|
"step": 31
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.7231638418079096,
|
||
|
|
"grad_norm": 1.3183085918426514,
|
||
|
|
"learning_rate": 5.5804645706261515e-06,
|
||
|
|
"loss": 1.1725330352783203,
|
||
|
|
"step": 32
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8135593220338984,
|
||
|
|
"grad_norm": 1.2263569831848145,
|
||
|
|
"learning_rate": 5.290724144552379e-06,
|
||
|
|
"loss": 1.1297430992126465,
|
||
|
|
"step": 33
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.903954802259887,
|
||
|
|
"grad_norm": 1.3964532613754272,
|
||
|
|
"learning_rate": 5e-06,
|
||
|
|
"loss": 1.2231621742248535,
|
||
|
|
"step": 34
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.994350282485876,
|
||
|
|
"grad_norm": 1.325929045677185,
|
||
|
|
"learning_rate": 4.7092758554476215e-06,
|
||
|
|
"loss": 1.1796724796295166,
|
||
|
|
"step": 35
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0,
|
||
|
|
"grad_norm": 3.460449695587158,
|
||
|
|
"learning_rate": 4.4195354293738484e-06,
|
||
|
|
"loss": 1.1036558151245117,
|
||
|
|
"step": 36
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0903954802259888,
|
||
|
|
"grad_norm": 1.3740507364273071,
|
||
|
|
"learning_rate": 4.131759111665349e-06,
|
||
|
|
"loss": 1.158125638961792,
|
||
|
|
"step": 37
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.1807909604519775,
|
||
|
|
"grad_norm": 1.2848109006881714,
|
||
|
|
"learning_rate": 3.8469206462878e-06,
|
||
|
|
"loss": 1.0927133560180664,
|
||
|
|
"step": 38
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.2711864406779663,
|
||
|
|
"grad_norm": 1.12660813331604,
|
||
|
|
"learning_rate": 3.5659838364445505e-06,
|
||
|
|
"loss": 1.0715622901916504,
|
||
|
|
"step": 39
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.361581920903955,
|
||
|
|
"grad_norm": 1.301928997039795,
|
||
|
|
"learning_rate": 3.289899283371657e-06,
|
||
|
|
"loss": 1.0645561218261719,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.4519774011299433,
|
||
|
|
"grad_norm": 1.1756473779678345,
|
||
|
|
"learning_rate": 3.019601169804216e-06,
|
||
|
|
"loss": 1.0882360935211182,
|
||
|
|
"step": 41
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.542372881355932,
|
||
|
|
"grad_norm": 1.3334362506866455,
|
||
|
|
"learning_rate": 2.7560040989976894e-06,
|
||
|
|
"loss": 1.096019983291626,
|
||
|
|
"step": 42
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.632768361581921,
|
||
|
|
"grad_norm": 1.1457642316818237,
|
||
|
|
"learning_rate": 2.5000000000000015e-06,
|
||
|
|
"loss": 1.1114509105682373,
|
||
|
|
"step": 43
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.7231638418079096,
|
||
|
|
"grad_norm": 1.169641137123108,
|
||
|
|
"learning_rate": 2.2524551096459703e-06,
|
||
|
|
"loss": 1.0899746417999268,
|
||
|
|
"step": 44
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.8135593220338984,
|
||
|
|
"grad_norm": 1.1258875131607056,
|
||
|
|
"learning_rate": 2.0142070414860704e-06,
|
||
|
|
"loss": 1.0599607229232788,
|
||
|
|
"step": 45
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.903954802259887,
|
||
|
|
"grad_norm": 1.2110517024993896,
|
||
|
|
"learning_rate": 1.7860619515673034e-06,
|
||
|
|
"loss": 1.045163869857788,
|
||
|
|
"step": 46
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.994350282485876,
|
||
|
|
"grad_norm": 1.1686667203903198,
|
||
|
|
"learning_rate": 1.5687918106563326e-06,
|
||
|
|
"loss": 1.0834063291549683,
|
||
|
|
"step": 47
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.0,
|
||
|
|
"grad_norm": 3.928598403930664,
|
||
|
|
"learning_rate": 1.3631317921347564e-06,
|
||
|
|
"loss": 0.9682188034057617,
|
||
|
|
"step": 48
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.090395480225989,
|
||
|
|
"grad_norm": 1.1086369752883911,
|
||
|
|
"learning_rate": 1.1697777844051105e-06,
|
||
|
|
"loss": 1.0497123003005981,
|
||
|
|
"step": 49
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.1807909604519775,
|
||
|
|
"grad_norm": 1.102365493774414,
|
||
|
|
"learning_rate": 9.893840362247809e-07,
|
||
|
|
"loss": 1.0524401664733887,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.271186440677966,
|
||
|
|
"grad_norm": 1.0737630128860474,
|
||
|
|
"learning_rate": 8.225609429353187e-07,
|
||
|
|
"loss": 1.0490361452102661,
|
||
|
|
"step": 51
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.361581920903955,
|
||
|
|
"grad_norm": 1.0565601587295532,
|
||
|
|
"learning_rate": 6.698729810778065e-07,
|
||
|
|
"loss": 1.036095380783081,
|
||
|
|
"step": 52
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.451977401129944,
|
||
|
|
"grad_norm": 0.9903903007507324,
|
||
|
|
"learning_rate": 5.318367983829393e-07,
|
||
|
|
"loss": 1.0308215618133545,
|
||
|
|
"step": 53
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.5423728813559325,
|
||
|
|
"grad_norm": 0.990847110748291,
|
||
|
|
"learning_rate": 4.089194655986306e-07,
|
||
|
|
"loss": 1.0118765830993652,
|
||
|
|
"step": 54
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.632768361581921,
|
||
|
|
"grad_norm": 1.103642463684082,
|
||
|
|
"learning_rate": 3.015368960704584e-07,
|
||
|
|
"loss": 1.0313677787780762,
|
||
|
|
"step": 55
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.72316384180791,
|
||
|
|
"grad_norm": 1.1910076141357422,
|
||
|
|
"learning_rate": 2.1005243842255552e-07,
|
||
|
|
"loss": 1.0596657991409302,
|
||
|
|
"step": 56
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.813559322033898,
|
||
|
|
"grad_norm": 0.9572083950042725,
|
||
|
|
"learning_rate": 1.3477564710088097e-07,
|
||
|
|
"loss": 1.0232012271881104,
|
||
|
|
"step": 57
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.903954802259887,
|
||
|
|
"grad_norm": 5.011074542999268,
|
||
|
|
"learning_rate": 7.59612349389599e-08,
|
||
|
|
"loss": 1.0278719663619995,
|
||
|
|
"step": 58
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 4.994350282485875,
|
||
|
|
"grad_norm": 1.3123799562454224,
|
||
|
|
"learning_rate": 3.3808211290284886e-08,
|
||
|
|
"loss": 1.0475084781646729,
|
||
|
|
"step": 59
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"grad_norm": 3.651301145553589,
|
||
|
|
"learning_rate": 8.459208643659122e-09,
|
||
|
|
"loss": 0.893081784248352,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 5.0,
|
||
|
|
"step": 60,
|
||
|
|
"total_flos": 6.322240708254106e+17,
|
||
|
|
"train_loss": 1.4024348258972168,
|
||
|
|
"train_runtime": 1552.6736,
|
||
|
|
"train_samples_per_second": 2.273,
|
||
|
|
"train_steps_per_second": 0.039
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 1,
|
||
|
|
"max_steps": 60,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 5,
|
||
|
|
"save_steps": 500,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 6.322240708254106e+17,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|