|
{ |
|
"best_metric": 0.3127128779888153, |
|
"best_model_checkpoint": "./convnext-base-8e-5-15ep/checkpoint-8792", |
|
"epoch": 15.0, |
|
"eval_steps": 500, |
|
"global_step": 16485, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.09, |
|
"grad_norm": 11.724858283996582, |
|
"learning_rate": 7.999273662119933e-05, |
|
"loss": 1.9321, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"grad_norm": 21.20332908630371, |
|
"learning_rate": 7.99709491226309e-05, |
|
"loss": 0.9556, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"grad_norm": 15.31360149383545, |
|
"learning_rate": 7.993464541683746e-05, |
|
"loss": 0.7852, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"grad_norm": 18.61130714416504, |
|
"learning_rate": 7.988383868819736e-05, |
|
"loss": 0.6454, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 20.7940616607666, |
|
"learning_rate": 7.981854738813642e-05, |
|
"loss": 0.6828, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 16.246265411376953, |
|
"learning_rate": 7.973879522842682e-05, |
|
"loss": 0.5941, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 14.500826835632324, |
|
"learning_rate": 7.964461117257593e-05, |
|
"loss": 0.6027, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 11.483868598937988, |
|
"learning_rate": 7.953602942530744e-05, |
|
"loss": 0.576, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"grad_norm": 16.98451805114746, |
|
"learning_rate": 7.941308942013942e-05, |
|
"loss": 0.5384, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"grad_norm": 6.033311367034912, |
|
"learning_rate": 7.927583580506324e-05, |
|
"loss": 0.5403, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.8743538767395627, |
|
"eval_loss": 0.42309147119522095, |
|
"eval_runtime": 106.7517, |
|
"eval_samples_per_second": 23.559, |
|
"eval_steps_per_second": 1.48, |
|
"step": 1099 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 15.62583065032959, |
|
"learning_rate": 7.91243184263288e-05, |
|
"loss": 0.5847, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 1.09, |
|
"grad_norm": 6.581829071044922, |
|
"learning_rate": 7.895859231034193e-05, |
|
"loss": 0.4455, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.18, |
|
"grad_norm": 12.816280364990234, |
|
"learning_rate": 7.877871764368051e-05, |
|
"loss": 0.4448, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 1.27, |
|
"grad_norm": 9.03258991241455, |
|
"learning_rate": 7.858475975123658e-05, |
|
"loss": 0.4507, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 1.36, |
|
"grad_norm": 17.637292861938477, |
|
"learning_rate": 7.837678907249235e-05, |
|
"loss": 0.405, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 1.46, |
|
"grad_norm": 11.983306884765625, |
|
"learning_rate": 7.815488113593875e-05, |
|
"loss": 0.3791, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.55, |
|
"grad_norm": 12.843159675598145, |
|
"learning_rate": 7.79191165316459e-05, |
|
"loss": 0.462, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 1.64, |
|
"grad_norm": 19.013002395629883, |
|
"learning_rate": 7.766958088199526e-05, |
|
"loss": 0.4481, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 1.73, |
|
"grad_norm": 5.846986770629883, |
|
"learning_rate": 7.740636481058419e-05, |
|
"loss": 0.4349, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 1.82, |
|
"grad_norm": 12.305352210998535, |
|
"learning_rate": 7.712956390931434e-05, |
|
"loss": 0.3673, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 1.91, |
|
"grad_norm": 8.298523902893066, |
|
"learning_rate": 7.683927870367564e-05, |
|
"loss": 0.4376, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.8898608349900596, |
|
"eval_loss": 0.38170379400253296, |
|
"eval_runtime": 106.7439, |
|
"eval_samples_per_second": 23.561, |
|
"eval_steps_per_second": 1.48, |
|
"step": 2198 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 14.847254753112793, |
|
"learning_rate": 7.653561461623853e-05, |
|
"loss": 0.4081, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 2.09, |
|
"grad_norm": 10.653617858886719, |
|
"learning_rate": 7.621868192836776e-05, |
|
"loss": 0.3557, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 2.18, |
|
"grad_norm": 16.83597755432129, |
|
"learning_rate": 7.588859574017165e-05, |
|
"loss": 0.3405, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 2.27, |
|
"grad_norm": 15.18385124206543, |
|
"learning_rate": 7.554547592870128e-05, |
|
"loss": 0.3196, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 2.37, |
|
"grad_norm": 5.773273468017578, |
|
"learning_rate": 7.518944710441491e-05, |
|
"loss": 0.37, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 2.46, |
|
"grad_norm": 13.849283218383789, |
|
"learning_rate": 7.482063856592323e-05, |
|
"loss": 0.3626, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 2.55, |
|
"grad_norm": 15.644866943359375, |
|
"learning_rate": 7.44391842530323e-05, |
|
"loss": 0.3598, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 2.64, |
|
"grad_norm": 11.187522888183594, |
|
"learning_rate": 7.404522269810056e-05, |
|
"loss": 0.3805, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 2.73, |
|
"grad_norm": 2.7877767086029053, |
|
"learning_rate": 7.363889697572835e-05, |
|
"loss": 0.322, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 2.82, |
|
"grad_norm": 11.111302375793457, |
|
"learning_rate": 7.322035465079756e-05, |
|
"loss": 0.3524, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 2.91, |
|
"grad_norm": 15.361136436462402, |
|
"learning_rate": 7.278974772488071e-05, |
|
"loss": 0.3601, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.9017892644135189, |
|
"eval_loss": 0.38179874420166016, |
|
"eval_runtime": 107.7627, |
|
"eval_samples_per_second": 23.338, |
|
"eval_steps_per_second": 1.466, |
|
"step": 3297 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"grad_norm": 11.831352233886719, |
|
"learning_rate": 7.234723258103863e-05, |
|
"loss": 0.3396, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 3.09, |
|
"grad_norm": 11.50969123840332, |
|
"learning_rate": 7.189296992702708e-05, |
|
"loss": 0.2502, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 3.18, |
|
"grad_norm": 0.8365811705589294, |
|
"learning_rate": 7.142712473693261e-05, |
|
"loss": 0.2747, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 3.28, |
|
"grad_norm": 11.02737045288086, |
|
"learning_rate": 7.094986619125911e-05, |
|
"loss": 0.293, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 3.37, |
|
"grad_norm": 8.201223373413086, |
|
"learning_rate": 7.046136761548677e-05, |
|
"loss": 0.2775, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 3.46, |
|
"grad_norm": 4.118073463439941, |
|
"learning_rate": 6.996180641712553e-05, |
|
"loss": 0.2622, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 3.55, |
|
"grad_norm": 16.599191665649414, |
|
"learning_rate": 6.945136402128628e-05, |
|
"loss": 0.3098, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 3.64, |
|
"grad_norm": 5.20058012008667, |
|
"learning_rate": 6.893022580479289e-05, |
|
"loss": 0.3156, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 3.73, |
|
"grad_norm": 7.4321465492248535, |
|
"learning_rate": 6.839858102885904e-05, |
|
"loss": 0.3304, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 3.82, |
|
"grad_norm": 0.302477091550827, |
|
"learning_rate": 6.785662277035447e-05, |
|
"loss": 0.2812, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 3.91, |
|
"grad_norm": 9.081244468688965, |
|
"learning_rate": 6.730454785168549e-05, |
|
"loss": 0.3202, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.920079522862823, |
|
"eval_loss": 0.3152866065502167, |
|
"eval_runtime": 107.0627, |
|
"eval_samples_per_second": 23.491, |
|
"eval_steps_per_second": 1.476, |
|
"step": 4396 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 7.102989673614502, |
|
"learning_rate": 6.67425567693151e-05, |
|
"loss": 0.2809, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 4.09, |
|
"grad_norm": 7.009576320648193, |
|
"learning_rate": 6.617085362094902e-05, |
|
"loss": 0.2443, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 4.19, |
|
"grad_norm": 7.170097827911377, |
|
"learning_rate": 6.558964603141364e-05, |
|
"loss": 0.2592, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 4.28, |
|
"grad_norm": 2.704582452774048, |
|
"learning_rate": 6.49991450772532e-05, |
|
"loss": 0.2488, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 4.37, |
|
"grad_norm": 0.38259217143058777, |
|
"learning_rate": 6.43995652100733e-05, |
|
"loss": 0.2191, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 4.46, |
|
"grad_norm": 11.570588111877441, |
|
"learning_rate": 6.379112417865878e-05, |
|
"loss": 0.2352, |
|
"step": 4900 |
|
}, |
|
{ |
|
"epoch": 4.55, |
|
"grad_norm": 14.971210479736328, |
|
"learning_rate": 6.317404294989409e-05, |
|
"loss": 0.2137, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 4.64, |
|
"grad_norm": 18.02596664428711, |
|
"learning_rate": 6.2548545628515e-05, |
|
"loss": 0.2284, |
|
"step": 5100 |
|
}, |
|
{ |
|
"epoch": 4.73, |
|
"grad_norm": 4.175428867340088, |
|
"learning_rate": 6.191485937572068e-05, |
|
"loss": 0.2541, |
|
"step": 5200 |
|
}, |
|
{ |
|
"epoch": 4.82, |
|
"grad_norm": 18.096067428588867, |
|
"learning_rate": 6.127321432667592e-05, |
|
"loss": 0.2382, |
|
"step": 5300 |
|
}, |
|
{ |
|
"epoch": 4.91, |
|
"grad_norm": 13.15633773803711, |
|
"learning_rate": 6.062384350693302e-05, |
|
"loss": 0.2463, |
|
"step": 5400 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.9228628230616303, |
|
"eval_loss": 0.32158657908439636, |
|
"eval_runtime": 107.3379, |
|
"eval_samples_per_second": 23.431, |
|
"eval_steps_per_second": 1.472, |
|
"step": 5495 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"grad_norm": 6.101550579071045, |
|
"learning_rate": 5.9966982747804285e-05, |
|
"loss": 0.2692, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 5.1, |
|
"grad_norm": 1.3759921789169312, |
|
"learning_rate": 5.930287060071538e-05, |
|
"loss": 0.1879, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 5.19, |
|
"grad_norm": 23.739288330078125, |
|
"learning_rate": 5.86317482505708e-05, |
|
"loss": 0.2182, |
|
"step": 5700 |
|
}, |
|
{ |
|
"epoch": 5.28, |
|
"grad_norm": 5.8365159034729, |
|
"learning_rate": 5.795385942816307e-05, |
|
"loss": 0.189, |
|
"step": 5800 |
|
}, |
|
{ |
|
"epoch": 5.37, |
|
"grad_norm": 0.29634252190589905, |
|
"learning_rate": 5.726945032165731e-05, |
|
"loss": 0.214, |
|
"step": 5900 |
|
}, |
|
{ |
|
"epoch": 5.46, |
|
"grad_norm": 0.2643405497074127, |
|
"learning_rate": 5.657876948718328e-05, |
|
"loss": 0.1798, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 5.55, |
|
"grad_norm": 0.6693936586380005, |
|
"learning_rate": 5.588206775856751e-05, |
|
"loss": 0.2094, |
|
"step": 6100 |
|
}, |
|
{ |
|
"epoch": 5.64, |
|
"grad_norm": 2.3573074340820312, |
|
"learning_rate": 5.517959815623834e-05, |
|
"loss": 0.1737, |
|
"step": 6200 |
|
}, |
|
{ |
|
"epoch": 5.73, |
|
"grad_norm": 6.905389308929443, |
|
"learning_rate": 5.447161579533662e-05, |
|
"loss": 0.2354, |
|
"step": 6300 |
|
}, |
|
{ |
|
"epoch": 5.82, |
|
"grad_norm": 5.905348300933838, |
|
"learning_rate": 5.375837779306594e-05, |
|
"loss": 0.2056, |
|
"step": 6400 |
|
}, |
|
{ |
|
"epoch": 5.91, |
|
"grad_norm": 13.43565559387207, |
|
"learning_rate": 5.3040143175315564e-05, |
|
"loss": 0.2051, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.921272365805169, |
|
"eval_loss": 0.33380183577537537, |
|
"eval_runtime": 106.6505, |
|
"eval_samples_per_second": 23.582, |
|
"eval_steps_per_second": 1.481, |
|
"step": 6594 |
|
}, |
|
{ |
|
"epoch": 6.01, |
|
"grad_norm": 8.566280364990234, |
|
"learning_rate": 5.2317172782590326e-05, |
|
"loss": 0.2282, |
|
"step": 6600 |
|
}, |
|
{ |
|
"epoch": 6.1, |
|
"grad_norm": 14.734358787536621, |
|
"learning_rate": 5.158972917528144e-05, |
|
"loss": 0.1602, |
|
"step": 6700 |
|
}, |
|
{ |
|
"epoch": 6.19, |
|
"grad_norm": 20.460796356201172, |
|
"learning_rate": 5.085807653831266e-05, |
|
"loss": 0.1738, |
|
"step": 6800 |
|
}, |
|
{ |
|
"epoch": 6.28, |
|
"grad_norm": 1.8641005754470825, |
|
"learning_rate": 5.012248058519667e-05, |
|
"loss": 0.1813, |
|
"step": 6900 |
|
}, |
|
{ |
|
"epoch": 6.37, |
|
"grad_norm": 0.38667044043540955, |
|
"learning_rate": 4.9383208461536056e-05, |
|
"loss": 0.1964, |
|
"step": 7000 |
|
}, |
|
{ |
|
"epoch": 6.46, |
|
"grad_norm": 8.61457347869873, |
|
"learning_rate": 4.8640528648004334e-05, |
|
"loss": 0.1706, |
|
"step": 7100 |
|
}, |
|
{ |
|
"epoch": 6.55, |
|
"grad_norm": 5.7172160148620605, |
|
"learning_rate": 4.789471086284219e-05, |
|
"loss": 0.1612, |
|
"step": 7200 |
|
}, |
|
{ |
|
"epoch": 6.64, |
|
"grad_norm": 0.22469474375247955, |
|
"learning_rate": 4.714602596390412e-05, |
|
"loss": 0.1697, |
|
"step": 7300 |
|
}, |
|
{ |
|
"epoch": 6.73, |
|
"grad_norm": 15.02285385131836, |
|
"learning_rate": 4.639474585029129e-05, |
|
"loss": 0.1886, |
|
"step": 7400 |
|
}, |
|
{ |
|
"epoch": 6.82, |
|
"grad_norm": 12.13104248046875, |
|
"learning_rate": 4.5641143363606216e-05, |
|
"loss": 0.1735, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 6.92, |
|
"grad_norm": 2.0161709785461426, |
|
"learning_rate": 4.488549218886521e-05, |
|
"loss": 0.2108, |
|
"step": 7600 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.9272365805168986, |
|
"eval_loss": 0.3223852515220642, |
|
"eval_runtime": 107.0584, |
|
"eval_samples_per_second": 23.492, |
|
"eval_steps_per_second": 1.476, |
|
"step": 7693 |
|
}, |
|
{ |
|
"epoch": 7.01, |
|
"grad_norm": 8.913517951965332, |
|
"learning_rate": 4.412806675510442e-05, |
|
"loss": 0.1466, |
|
"step": 7700 |
|
}, |
|
{ |
|
"epoch": 7.1, |
|
"grad_norm": 9.739709854125977, |
|
"learning_rate": 4.33691421357158e-05, |
|
"loss": 0.1382, |
|
"step": 7800 |
|
}, |
|
{ |
|
"epoch": 7.19, |
|
"grad_norm": 0.3633786737918854, |
|
"learning_rate": 4.260899394854893e-05, |
|
"loss": 0.1598, |
|
"step": 7900 |
|
}, |
|
{ |
|
"epoch": 7.28, |
|
"grad_norm": 1.9761762619018555, |
|
"learning_rate": 4.184789825581521e-05, |
|
"loss": 0.1272, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 7.37, |
|
"grad_norm": 0.9676409363746643, |
|
"learning_rate": 4.108613146383063e-05, |
|
"loss": 0.1692, |
|
"step": 8100 |
|
}, |
|
{ |
|
"epoch": 7.46, |
|
"grad_norm": 0.3291204273700714, |
|
"learning_rate": 4.03239702226336e-05, |
|
"loss": 0.1301, |
|
"step": 8200 |
|
}, |
|
{ |
|
"epoch": 7.55, |
|
"grad_norm": 11.293334007263184, |
|
"learning_rate": 3.9561691325514226e-05, |
|
"loss": 0.1496, |
|
"step": 8300 |
|
}, |
|
{ |
|
"epoch": 7.64, |
|
"grad_norm": 0.4240261912345886, |
|
"learning_rate": 3.879957160849155e-05, |
|
"loss": 0.1658, |
|
"step": 8400 |
|
}, |
|
{ |
|
"epoch": 7.73, |
|
"grad_norm": 9.459733009338379, |
|
"learning_rate": 3.803788784977541e-05, |
|
"loss": 0.1848, |
|
"step": 8500 |
|
}, |
|
{ |
|
"epoch": 7.83, |
|
"grad_norm": 0.03799422085285187, |
|
"learning_rate": 3.7276916669249106e-05, |
|
"loss": 0.1599, |
|
"step": 8600 |
|
}, |
|
{ |
|
"epoch": 7.92, |
|
"grad_norm": 7.809151649475098, |
|
"learning_rate": 3.651693442800964e-05, |
|
"loss": 0.1568, |
|
"step": 8700 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.9355864811133201, |
|
"eval_loss": 0.3127128779888153, |
|
"eval_runtime": 107.9901, |
|
"eval_samples_per_second": 23.289, |
|
"eval_steps_per_second": 1.463, |
|
"step": 8792 |
|
}, |
|
{ |
|
"epoch": 8.01, |
|
"grad_norm": 0.09405653923749924, |
|
"learning_rate": 3.575821712800204e-05, |
|
"loss": 0.1238, |
|
"step": 8800 |
|
}, |
|
{ |
|
"epoch": 8.1, |
|
"grad_norm": 5.202500820159912, |
|
"learning_rate": 3.5001040311783905e-05, |
|
"loss": 0.1327, |
|
"step": 8900 |
|
}, |
|
{ |
|
"epoch": 8.19, |
|
"grad_norm": 0.39644354581832886, |
|
"learning_rate": 3.4245678962457005e-05, |
|
"loss": 0.1265, |
|
"step": 9000 |
|
}, |
|
{ |
|
"epoch": 8.28, |
|
"grad_norm": 3.383341073989868, |
|
"learning_rate": 3.349240740380192e-05, |
|
"loss": 0.1159, |
|
"step": 9100 |
|
}, |
|
{ |
|
"epoch": 8.37, |
|
"grad_norm": 0.2011689841747284, |
|
"learning_rate": 3.274149920065217e-05, |
|
"loss": 0.1529, |
|
"step": 9200 |
|
}, |
|
{ |
|
"epoch": 8.46, |
|
"grad_norm": 0.02637963555753231, |
|
"learning_rate": 3.199322705954396e-05, |
|
"loss": 0.1003, |
|
"step": 9300 |
|
}, |
|
{ |
|
"epoch": 8.55, |
|
"grad_norm": 6.775584697723389, |
|
"learning_rate": 3.1247862729677615e-05, |
|
"loss": 0.1186, |
|
"step": 9400 |
|
}, |
|
{ |
|
"epoch": 8.64, |
|
"grad_norm": 0.8526474833488464, |
|
"learning_rate": 3.0505676904226766e-05, |
|
"loss": 0.1354, |
|
"step": 9500 |
|
}, |
|
{ |
|
"epoch": 8.74, |
|
"grad_norm": 5.6588945388793945, |
|
"learning_rate": 2.9766939122030964e-05, |
|
"loss": 0.137, |
|
"step": 9600 |
|
}, |
|
{ |
|
"epoch": 8.83, |
|
"grad_norm": 13.947266578674316, |
|
"learning_rate": 2.903191766970749e-05, |
|
"loss": 0.1336, |
|
"step": 9700 |
|
}, |
|
{ |
|
"epoch": 8.92, |
|
"grad_norm": 6.7341694831848145, |
|
"learning_rate": 2.8300879484218112e-05, |
|
"loss": 0.1491, |
|
"step": 9800 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.9351888667992048, |
|
"eval_loss": 0.32577136158943176, |
|
"eval_runtime": 108.1206, |
|
"eval_samples_per_second": 23.261, |
|
"eval_steps_per_second": 1.461, |
|
"step": 9891 |
|
}, |
|
{ |
|
"epoch": 9.01, |
|
"grad_norm": 1.849369764328003, |
|
"learning_rate": 2.757409005592578e-05, |
|
"loss": 0.1105, |
|
"step": 9900 |
|
}, |
|
{ |
|
"epoch": 9.1, |
|
"grad_norm": 0.0076730018481612206, |
|
"learning_rate": 2.6851813332176796e-05, |
|
"loss": 0.09, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 9.19, |
|
"grad_norm": 0.42202454805374146, |
|
"learning_rate": 2.6134311621443315e-05, |
|
"loss": 0.1067, |
|
"step": 10100 |
|
}, |
|
{ |
|
"epoch": 9.28, |
|
"grad_norm": 0.32351329922676086, |
|
"learning_rate": 2.5421845498061117e-05, |
|
"loss": 0.1117, |
|
"step": 10200 |
|
}, |
|
{ |
|
"epoch": 9.37, |
|
"grad_norm": 0.507881760597229, |
|
"learning_rate": 2.4714673707597038e-05, |
|
"loss": 0.099, |
|
"step": 10300 |
|
}, |
|
{ |
|
"epoch": 9.46, |
|
"grad_norm": 16.288957595825195, |
|
"learning_rate": 2.4013053072880646e-05, |
|
"loss": 0.091, |
|
"step": 10400 |
|
}, |
|
{ |
|
"epoch": 9.55, |
|
"grad_norm": 0.02297704666852951, |
|
"learning_rate": 2.3317238400734145e-05, |
|
"loss": 0.1295, |
|
"step": 10500 |
|
}, |
|
{ |
|
"epoch": 9.65, |
|
"grad_norm": 9.774455070495605, |
|
"learning_rate": 2.2627482389434477e-05, |
|
"loss": 0.1244, |
|
"step": 10600 |
|
}, |
|
{ |
|
"epoch": 9.74, |
|
"grad_norm": 1.1963157653808594, |
|
"learning_rate": 2.194403553694116e-05, |
|
"loss": 0.1319, |
|
"step": 10700 |
|
}, |
|
{ |
|
"epoch": 9.83, |
|
"grad_norm": 0.8269553184509277, |
|
"learning_rate": 2.1267146049923167e-05, |
|
"loss": 0.1165, |
|
"step": 10800 |
|
}, |
|
{ |
|
"epoch": 9.92, |
|
"grad_norm": 0.5792891383171082, |
|
"learning_rate": 2.059705975361802e-05, |
|
"loss": 0.0998, |
|
"step": 10900 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.9347912524850894, |
|
"eval_loss": 0.3257232904434204, |
|
"eval_runtime": 107.8035, |
|
"eval_samples_per_second": 23.329, |
|
"eval_steps_per_second": 1.466, |
|
"step": 10990 |
|
}, |
|
{ |
|
"epoch": 10.01, |
|
"grad_norm": 0.3180704414844513, |
|
"learning_rate": 1.9934020002555702e-05, |
|
"loss": 0.1086, |
|
"step": 11000 |
|
}, |
|
{ |
|
"epoch": 10.1, |
|
"grad_norm": 7.30010461807251, |
|
"learning_rate": 1.9278267592179803e-05, |
|
"loss": 0.0895, |
|
"step": 11100 |
|
}, |
|
{ |
|
"epoch": 10.19, |
|
"grad_norm": 0.1635340303182602, |
|
"learning_rate": 1.86300406713981e-05, |
|
"loss": 0.0777, |
|
"step": 11200 |
|
}, |
|
{ |
|
"epoch": 10.28, |
|
"grad_norm": 2.3367292881011963, |
|
"learning_rate": 1.7989574656094327e-05, |
|
"loss": 0.0737, |
|
"step": 11300 |
|
}, |
|
{ |
|
"epoch": 10.37, |
|
"grad_norm": 0.8518896698951721, |
|
"learning_rate": 1.7357102143632386e-05, |
|
"loss": 0.1099, |
|
"step": 11400 |
|
}, |
|
{ |
|
"epoch": 10.46, |
|
"grad_norm": 0.14773418009281158, |
|
"learning_rate": 1.6732852828384235e-05, |
|
"loss": 0.1096, |
|
"step": 11500 |
|
}, |
|
{ |
|
"epoch": 10.56, |
|
"grad_norm": 0.056224796921014786, |
|
"learning_rate": 1.6117053418311988e-05, |
|
"loss": 0.0624, |
|
"step": 11600 |
|
}, |
|
{ |
|
"epoch": 10.65, |
|
"grad_norm": 14.274983406066895, |
|
"learning_rate": 1.550992755263471e-05, |
|
"loss": 0.0981, |
|
"step": 11700 |
|
}, |
|
{ |
|
"epoch": 10.74, |
|
"grad_norm": 15.517501831054688, |
|
"learning_rate": 1.4911695720609486e-05, |
|
"loss": 0.1016, |
|
"step": 11800 |
|
}, |
|
{ |
|
"epoch": 10.83, |
|
"grad_norm": 1.103292465209961, |
|
"learning_rate": 1.4322575181456664e-05, |
|
"loss": 0.1021, |
|
"step": 11900 |
|
}, |
|
{ |
|
"epoch": 10.92, |
|
"grad_norm": 0.8106482625007629, |
|
"learning_rate": 1.3742779885457966e-05, |
|
"loss": 0.0728, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.936779324055666, |
|
"eval_loss": 0.3289387822151184, |
|
"eval_runtime": 108.2127, |
|
"eval_samples_per_second": 23.241, |
|
"eval_steps_per_second": 1.46, |
|
"step": 12089 |
|
}, |
|
{ |
|
"epoch": 11.01, |
|
"grad_norm": 0.03628486394882202, |
|
"learning_rate": 1.3172520396256503e-05, |
|
"loss": 0.0635, |
|
"step": 12100 |
|
}, |
|
{ |
|
"epoch": 11.1, |
|
"grad_norm": 0.15396057069301605, |
|
"learning_rate": 1.2612003814386489e-05, |
|
"loss": 0.0998, |
|
"step": 12200 |
|
}, |
|
{ |
|
"epoch": 11.19, |
|
"grad_norm": 0.37192150950431824, |
|
"learning_rate": 1.206143370206086e-05, |
|
"loss": 0.0794, |
|
"step": 12300 |
|
}, |
|
{ |
|
"epoch": 11.28, |
|
"grad_norm": 0.3847203254699707, |
|
"learning_rate": 1.1521010009243714e-05, |
|
"loss": 0.0638, |
|
"step": 12400 |
|
}, |
|
{ |
|
"epoch": 11.37, |
|
"grad_norm": 7.261880874633789, |
|
"learning_rate": 1.0990929001034738e-05, |
|
"loss": 0.0738, |
|
"step": 12500 |
|
}, |
|
{ |
|
"epoch": 11.46, |
|
"grad_norm": 0.17248640954494476, |
|
"learning_rate": 1.0471383186391817e-05, |
|
"loss": 0.0636, |
|
"step": 12600 |
|
}, |
|
{ |
|
"epoch": 11.56, |
|
"grad_norm": 0.031174590811133385, |
|
"learning_rate": 9.962561248217764e-06, |
|
"loss": 0.0914, |
|
"step": 12700 |
|
}, |
|
{ |
|
"epoch": 11.65, |
|
"grad_norm": 1.3612018823623657, |
|
"learning_rate": 9.464647974836496e-06, |
|
"loss": 0.0781, |
|
"step": 12800 |
|
}, |
|
{ |
|
"epoch": 11.74, |
|
"grad_norm": 0.09444615244865417, |
|
"learning_rate": 8.977824192883772e-06, |
|
"loss": 0.067, |
|
"step": 12900 |
|
}, |
|
{ |
|
"epoch": 11.83, |
|
"grad_norm": 7.895622253417969, |
|
"learning_rate": 8.502266701636457e-06, |
|
"loss": 0.0846, |
|
"step": 13000 |
|
}, |
|
{ |
|
"epoch": 11.92, |
|
"grad_norm": 0.03469648212194443, |
|
"learning_rate": 8.038148208804567e-06, |
|
"loss": 0.0872, |
|
"step": 13100 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.9407554671968191, |
|
"eval_loss": 0.32880690693855286, |
|
"eval_runtime": 107.2934, |
|
"eval_samples_per_second": 23.44, |
|
"eval_steps_per_second": 1.473, |
|
"step": 13188 |
|
}, |
|
{ |
|
"epoch": 12.01, |
|
"grad_norm": 0.054643452167510986, |
|
"learning_rate": 7.5856372678091964e-06, |
|
"loss": 0.0724, |
|
"step": 13200 |
|
}, |
|
{ |
|
"epoch": 12.1, |
|
"grad_norm": 14.980018615722656, |
|
"learning_rate": 7.14489821656914e-06, |
|
"loss": 0.0765, |
|
"step": 13300 |
|
}, |
|
{ |
|
"epoch": 12.19, |
|
"grad_norm": 0.03025561198592186, |
|
"learning_rate": 6.716091117818484e-06, |
|
"loss": 0.0754, |
|
"step": 13400 |
|
}, |
|
{ |
|
"epoch": 12.28, |
|
"grad_norm": 0.03179701790213585, |
|
"learning_rate": 6.29937170097672e-06, |
|
"loss": 0.0551, |
|
"step": 13500 |
|
}, |
|
{ |
|
"epoch": 12.37, |
|
"grad_norm": 8.448119163513184, |
|
"learning_rate": 5.8948913055927894e-06, |
|
"loss": 0.0952, |
|
"step": 13600 |
|
}, |
|
{ |
|
"epoch": 12.47, |
|
"grad_norm": 0.0018671215511858463, |
|
"learning_rate": 5.502796826383141e-06, |
|
"loss": 0.0713, |
|
"step": 13700 |
|
}, |
|
{ |
|
"epoch": 12.56, |
|
"grad_norm": 0.37293198704719543, |
|
"learning_rate": 5.123230659884178e-06, |
|
"loss": 0.063, |
|
"step": 13800 |
|
}, |
|
{ |
|
"epoch": 12.65, |
|
"grad_norm": 6.294044494628906, |
|
"learning_rate": 4.7563306527382615e-06, |
|
"loss": 0.0732, |
|
"step": 13900 |
|
}, |
|
{ |
|
"epoch": 12.74, |
|
"grad_norm": 8.385093688964844, |
|
"learning_rate": 4.402230051632081e-06, |
|
"loss": 0.0579, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 12.83, |
|
"grad_norm": 0.035636261105537415, |
|
"learning_rate": 4.061057454905615e-06, |
|
"loss": 0.1073, |
|
"step": 14100 |
|
}, |
|
{ |
|
"epoch": 12.92, |
|
"grad_norm": 9.956852912902832, |
|
"learning_rate": 3.7329367658491778e-06, |
|
"loss": 0.0731, |
|
"step": 14200 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.9407554671968191, |
|
"eval_loss": 0.3320312201976776, |
|
"eval_runtime": 107.2304, |
|
"eval_samples_per_second": 23.454, |
|
"eval_steps_per_second": 1.473, |
|
"step": 14287 |
|
}, |
|
{ |
|
"epoch": 13.01, |
|
"grad_norm": 0.09960483759641647, |
|
"learning_rate": 3.417987147705617e-06, |
|
"loss": 0.0629, |
|
"step": 14300 |
|
}, |
|
{ |
|
"epoch": 13.1, |
|
"grad_norm": 0.04828275740146637, |
|
"learning_rate": 3.116322980393922e-06, |
|
"loss": 0.0595, |
|
"step": 14400 |
|
}, |
|
{ |
|
"epoch": 13.19, |
|
"grad_norm": 7.161159515380859, |
|
"learning_rate": 2.828053818969978e-06, |
|
"loss": 0.092, |
|
"step": 14500 |
|
}, |
|
{ |
|
"epoch": 13.28, |
|
"grad_norm": 3.226339340209961, |
|
"learning_rate": 2.5532843538395868e-06, |
|
"loss": 0.074, |
|
"step": 14600 |
|
}, |
|
{ |
|
"epoch": 13.38, |
|
"grad_norm": 0.04475373029708862, |
|
"learning_rate": 2.2921143727381525e-06, |
|
"loss": 0.0591, |
|
"step": 14700 |
|
}, |
|
{ |
|
"epoch": 13.47, |
|
"grad_norm": 7.270404815673828, |
|
"learning_rate": 2.0446387244908774e-06, |
|
"loss": 0.0666, |
|
"step": 14800 |
|
}, |
|
{ |
|
"epoch": 13.56, |
|
"grad_norm": 0.04449067637324333, |
|
"learning_rate": 1.8109472845666198e-06, |
|
"loss": 0.0834, |
|
"step": 14900 |
|
}, |
|
{ |
|
"epoch": 13.65, |
|
"grad_norm": 20.532730102539062, |
|
"learning_rate": 1.5911249224379189e-06, |
|
"loss": 0.0676, |
|
"step": 15000 |
|
}, |
|
{ |
|
"epoch": 13.74, |
|
"grad_norm": 6.280622959136963, |
|
"learning_rate": 1.385251470759017e-06, |
|
"loss": 0.076, |
|
"step": 15100 |
|
}, |
|
{ |
|
"epoch": 13.83, |
|
"grad_norm": 0.3242604732513428, |
|
"learning_rate": 1.1934016963731465e-06, |
|
"loss": 0.0624, |
|
"step": 15200 |
|
}, |
|
{ |
|
"epoch": 13.92, |
|
"grad_norm": 5.6159210205078125, |
|
"learning_rate": 1.0156452731595112e-06, |
|
"loss": 0.0693, |
|
"step": 15300 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.9423459244532804, |
|
"eval_loss": 0.32697632908821106, |
|
"eval_runtime": 107.0584, |
|
"eval_samples_per_second": 23.492, |
|
"eval_steps_per_second": 1.476, |
|
"step": 15386 |
|
}, |
|
{ |
|
"epoch": 14.01, |
|
"grad_norm": 3.9605867862701416, |
|
"learning_rate": 8.520467567299184e-07, |
|
"loss": 0.0651, |
|
"step": 15400 |
|
}, |
|
{ |
|
"epoch": 14.1, |
|
"grad_norm": 0.07699358463287354, |
|
"learning_rate": 7.026655609841682e-07, |
|
"loss": 0.0655, |
|
"step": 15500 |
|
}, |
|
{ |
|
"epoch": 14.19, |
|
"grad_norm": 12.785538673400879, |
|
"learning_rate": 5.675559365327798e-07, |
|
"loss": 0.0736, |
|
"step": 15600 |
|
}, |
|
{ |
|
"epoch": 14.29, |
|
"grad_norm": 0.0682230219244957, |
|
"learning_rate": 4.467669509948591e-07, |
|
"loss": 0.0712, |
|
"step": 15700 |
|
}, |
|
{ |
|
"epoch": 14.38, |
|
"grad_norm": 5.338863849639893, |
|
"learning_rate": 3.403424711782455e-07, |
|
"loss": 0.076, |
|
"step": 15800 |
|
}, |
|
{ |
|
"epoch": 14.47, |
|
"grad_norm": 0.04831124097108841, |
|
"learning_rate": 2.4832114714847367e-07, |
|
"loss": 0.0593, |
|
"step": 15900 |
|
}, |
|
{ |
|
"epoch": 14.56, |
|
"grad_norm": 19.2628231048584, |
|
"learning_rate": 1.7073639819224608e-07, |
|
"loss": 0.0646, |
|
"step": 16000 |
|
}, |
|
{ |
|
"epoch": 14.65, |
|
"grad_norm": 0.4399223327636719, |
|
"learning_rate": 1.0761640068060353e-07, |
|
"loss": 0.0554, |
|
"step": 16100 |
|
}, |
|
{ |
|
"epoch": 14.74, |
|
"grad_norm": 15.196357727050781, |
|
"learning_rate": 5.898407783614169e-08, |
|
"loss": 0.0849, |
|
"step": 16200 |
|
}, |
|
{ |
|
"epoch": 14.83, |
|
"grad_norm": 0.01677566207945347, |
|
"learning_rate": 2.48570914079993e-08, |
|
"loss": 0.0743, |
|
"step": 16300 |
|
}, |
|
{ |
|
"epoch": 14.92, |
|
"grad_norm": 3.5998127460479736, |
|
"learning_rate": 5.247835257664591e-09, |
|
"loss": 0.0586, |
|
"step": 16400 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.9427435387673956, |
|
"eval_loss": 0.3250676989555359, |
|
"eval_runtime": 107.5456, |
|
"eval_samples_per_second": 23.385, |
|
"eval_steps_per_second": 1.469, |
|
"step": 16485 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"step": 16485, |
|
"total_flos": 6.140249030814106e+19, |
|
"train_loss": 0.21494813935844184, |
|
"train_runtime": 26072.0492, |
|
"train_samples_per_second": 10.115, |
|
"train_steps_per_second": 0.632 |
|
} |
|
], |
|
"logging_steps": 100, |
|
"max_steps": 16485, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 15, |
|
"save_steps": 500, |
|
"total_flos": 6.140249030814106e+19, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|