|
{ |
|
"best_metric": 0.7498265504837036, |
|
"best_model_checkpoint": "Action_model/checkpoint-900", |
|
"epoch": 10.0, |
|
"eval_steps": 100, |
|
"global_step": 6280, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 1.9456478357315063, |
|
"learning_rate": 9.968152866242038e-05, |
|
"loss": 2.2288, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"grad_norm": 2.081900119781494, |
|
"learning_rate": 9.936305732484077e-05, |
|
"loss": 2.0145, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 2.330000877380371, |
|
"learning_rate": 9.904458598726115e-05, |
|
"loss": 1.8019, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"grad_norm": 3.2511394023895264, |
|
"learning_rate": 9.874203821656052e-05, |
|
"loss": 1.5067, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"grad_norm": 2.664090394973755, |
|
"learning_rate": 9.842356687898089e-05, |
|
"loss": 1.3203, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"eval_accuracy": 0.7447619047619047, |
|
"eval_loss": 1.1809700727462769, |
|
"eval_runtime": 19.4439, |
|
"eval_samples_per_second": 54.001, |
|
"eval_steps_per_second": 6.789, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"grad_norm": 2.723193407058716, |
|
"learning_rate": 9.810509554140128e-05, |
|
"loss": 1.2276, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"grad_norm": 3.213502883911133, |
|
"learning_rate": 9.778662420382166e-05, |
|
"loss": 1.1515, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"grad_norm": 3.7674548625946045, |
|
"learning_rate": 9.746815286624204e-05, |
|
"loss": 1.0409, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"grad_norm": 2.553504228591919, |
|
"learning_rate": 9.714968152866243e-05, |
|
"loss": 1.016, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"grad_norm": 4.68205451965332, |
|
"learning_rate": 9.683121019108281e-05, |
|
"loss": 0.998, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"eval_accuracy": 0.7704761904761904, |
|
"eval_loss": 0.8924565315246582, |
|
"eval_runtime": 15.156, |
|
"eval_samples_per_second": 69.28, |
|
"eval_steps_per_second": 8.709, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 3.2618489265441895, |
|
"learning_rate": 9.651273885350319e-05, |
|
"loss": 0.984, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 2.7326881885528564, |
|
"learning_rate": 9.619426751592358e-05, |
|
"loss": 0.9017, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"grad_norm": 4.338141918182373, |
|
"learning_rate": 9.587579617834396e-05, |
|
"loss": 0.8816, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 5.7808918952941895, |
|
"learning_rate": 9.555732484076433e-05, |
|
"loss": 0.7619, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 5.106058120727539, |
|
"learning_rate": 9.523885350318473e-05, |
|
"loss": 0.765, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"eval_accuracy": 0.7638095238095238, |
|
"eval_loss": 0.8198639750480652, |
|
"eval_runtime": 15.6032, |
|
"eval_samples_per_second": 67.294, |
|
"eval_steps_per_second": 8.46, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"grad_norm": 2.242779493331909, |
|
"learning_rate": 9.49203821656051e-05, |
|
"loss": 0.8391, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"grad_norm": 5.80740213394165, |
|
"learning_rate": 9.460191082802548e-05, |
|
"loss": 0.6795, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"grad_norm": 5.4263153076171875, |
|
"learning_rate": 9.428343949044587e-05, |
|
"loss": 0.7903, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"grad_norm": 3.731250762939453, |
|
"learning_rate": 9.396496815286625e-05, |
|
"loss": 0.7099, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"grad_norm": 5.383817672729492, |
|
"learning_rate": 9.364649681528663e-05, |
|
"loss": 0.6521, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"eval_accuracy": 0.7371428571428571, |
|
"eval_loss": 0.8275942206382751, |
|
"eval_runtime": 17.2731, |
|
"eval_samples_per_second": 60.788, |
|
"eval_steps_per_second": 7.642, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"grad_norm": 4.755931377410889, |
|
"learning_rate": 9.332802547770702e-05, |
|
"loss": 0.6679, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"grad_norm": 3.101043462753296, |
|
"learning_rate": 9.30095541401274e-05, |
|
"loss": 0.5997, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 3.172748327255249, |
|
"learning_rate": 9.269108280254777e-05, |
|
"loss": 0.738, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 2.2239062786102295, |
|
"learning_rate": 9.237261146496817e-05, |
|
"loss": 0.5563, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 4.797071933746338, |
|
"learning_rate": 9.205414012738854e-05, |
|
"loss": 0.7612, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"eval_accuracy": 0.7209523809523809, |
|
"eval_loss": 0.8631104230880737, |
|
"eval_runtime": 15.8902, |
|
"eval_samples_per_second": 66.078, |
|
"eval_steps_per_second": 8.307, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 6.720874786376953, |
|
"learning_rate": 9.173566878980892e-05, |
|
"loss": 0.6968, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"grad_norm": 2.4448862075805664, |
|
"learning_rate": 9.141719745222931e-05, |
|
"loss": 0.5712, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"grad_norm": 4.25390100479126, |
|
"learning_rate": 9.109872611464969e-05, |
|
"loss": 0.6566, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"grad_norm": 4.468232154846191, |
|
"learning_rate": 9.078025477707007e-05, |
|
"loss": 0.6277, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"grad_norm": 5.054867267608643, |
|
"learning_rate": 9.046178343949046e-05, |
|
"loss": 0.5894, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"eval_accuracy": 0.7476190476190476, |
|
"eval_loss": 0.8175568580627441, |
|
"eval_runtime": 18.194, |
|
"eval_samples_per_second": 57.711, |
|
"eval_steps_per_second": 7.255, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"grad_norm": 3.530838966369629, |
|
"learning_rate": 9.014331210191084e-05, |
|
"loss": 0.5674, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 1.02, |
|
"grad_norm": 2.4299445152282715, |
|
"learning_rate": 8.982484076433122e-05, |
|
"loss": 0.5017, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 1.05, |
|
"grad_norm": 9.271321296691895, |
|
"learning_rate": 8.950636942675161e-05, |
|
"loss": 0.4916, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 1.08, |
|
"grad_norm": 3.6438238620758057, |
|
"learning_rate": 8.918789808917198e-05, |
|
"loss": 0.5967, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"grad_norm": 11.338240623474121, |
|
"learning_rate": 8.886942675159236e-05, |
|
"loss": 0.5381, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"eval_accuracy": 0.7523809523809524, |
|
"eval_loss": 0.7964714765548706, |
|
"eval_runtime": 16.275, |
|
"eval_samples_per_second": 64.516, |
|
"eval_steps_per_second": 8.111, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 1.15, |
|
"grad_norm": 7.793953895568848, |
|
"learning_rate": 8.855095541401275e-05, |
|
"loss": 0.4818, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 1.18, |
|
"grad_norm": 3.596195697784424, |
|
"learning_rate": 8.823248407643313e-05, |
|
"loss": 0.5585, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 1.21, |
|
"grad_norm": 2.5034844875335693, |
|
"learning_rate": 8.791401273885351e-05, |
|
"loss": 0.4471, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 1.24, |
|
"grad_norm": 2.8802740573883057, |
|
"learning_rate": 8.75955414012739e-05, |
|
"loss": 0.4908, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 1.27, |
|
"grad_norm": 3.318392753601074, |
|
"learning_rate": 8.727707006369426e-05, |
|
"loss": 0.4066, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 1.27, |
|
"eval_accuracy": 0.7485714285714286, |
|
"eval_loss": 0.8102111220359802, |
|
"eval_runtime": 15.1922, |
|
"eval_samples_per_second": 69.115, |
|
"eval_steps_per_second": 8.689, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 1.31, |
|
"grad_norm": 0.7513989806175232, |
|
"learning_rate": 8.695859872611464e-05, |
|
"loss": 0.3771, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 1.34, |
|
"grad_norm": 2.7686688899993896, |
|
"learning_rate": 8.664012738853503e-05, |
|
"loss": 0.4788, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 1.37, |
|
"grad_norm": 7.201110363006592, |
|
"learning_rate": 8.632165605095541e-05, |
|
"loss": 0.4752, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 1.4, |
|
"grad_norm": 5.174149513244629, |
|
"learning_rate": 8.600318471337579e-05, |
|
"loss": 0.4771, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 1.43, |
|
"grad_norm": 2.768077850341797, |
|
"learning_rate": 8.568471337579618e-05, |
|
"loss": 0.4825, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 1.43, |
|
"eval_accuracy": 0.7742857142857142, |
|
"eval_loss": 0.7498265504837036, |
|
"eval_runtime": 15.2211, |
|
"eval_samples_per_second": 68.983, |
|
"eval_steps_per_second": 8.672, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 1.46, |
|
"grad_norm": 5.662777423858643, |
|
"learning_rate": 8.536624203821656e-05, |
|
"loss": 0.3955, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 1.5, |
|
"grad_norm": 2.15919828414917, |
|
"learning_rate": 8.504777070063694e-05, |
|
"loss": 0.3685, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 1.53, |
|
"grad_norm": 7.603888988494873, |
|
"learning_rate": 8.472929936305733e-05, |
|
"loss": 0.3679, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 1.56, |
|
"grad_norm": 3.6407835483551025, |
|
"learning_rate": 8.44108280254777e-05, |
|
"loss": 0.4757, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 1.59, |
|
"grad_norm": 3.8537333011627197, |
|
"learning_rate": 8.409235668789808e-05, |
|
"loss": 0.4955, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 1.59, |
|
"eval_accuracy": 0.7019047619047619, |
|
"eval_loss": 0.9751647710800171, |
|
"eval_runtime": 15.4284, |
|
"eval_samples_per_second": 68.056, |
|
"eval_steps_per_second": 8.556, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 1.62, |
|
"grad_norm": 3.7005233764648438, |
|
"learning_rate": 8.377388535031847e-05, |
|
"loss": 0.4366, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 1.66, |
|
"grad_norm": 4.745129585266113, |
|
"learning_rate": 8.345541401273885e-05, |
|
"loss": 0.4014, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 1.69, |
|
"grad_norm": 2.9663562774658203, |
|
"learning_rate": 8.313694267515923e-05, |
|
"loss": 0.4432, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 1.72, |
|
"grad_norm": 3.435357093811035, |
|
"learning_rate": 8.281847133757962e-05, |
|
"loss": 0.475, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 1.75, |
|
"grad_norm": 8.792357444763184, |
|
"learning_rate": 8.25e-05, |
|
"loss": 0.3945, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 1.75, |
|
"eval_accuracy": 0.7380952380952381, |
|
"eval_loss": 0.8149927854537964, |
|
"eval_runtime": 15.9868, |
|
"eval_samples_per_second": 65.679, |
|
"eval_steps_per_second": 8.257, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 1.78, |
|
"grad_norm": 4.0536346435546875, |
|
"learning_rate": 8.218152866242038e-05, |
|
"loss": 0.4341, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 1.82, |
|
"grad_norm": 4.5750732421875, |
|
"learning_rate": 8.186305732484077e-05, |
|
"loss": 0.2879, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 1.85, |
|
"grad_norm": 9.274312019348145, |
|
"learning_rate": 8.154458598726115e-05, |
|
"loss": 0.4391, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 1.88, |
|
"grad_norm": 3.804741859436035, |
|
"learning_rate": 8.122611464968152e-05, |
|
"loss": 0.4918, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 1.91, |
|
"grad_norm": 8.260722160339355, |
|
"learning_rate": 8.090764331210192e-05, |
|
"loss": 0.4142, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.91, |
|
"eval_accuracy": 0.7609523809523809, |
|
"eval_loss": 0.7953311204910278, |
|
"eval_runtime": 15.4376, |
|
"eval_samples_per_second": 68.016, |
|
"eval_steps_per_second": 8.551, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.94, |
|
"grad_norm": 6.791586399078369, |
|
"learning_rate": 8.05891719745223e-05, |
|
"loss": 0.3653, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 1.97, |
|
"grad_norm": 5.216760158538818, |
|
"learning_rate": 8.027070063694267e-05, |
|
"loss": 0.3321, |
|
"step": 1240 |
|
}, |
|
{ |
|
"epoch": 2.01, |
|
"grad_norm": 2.542558431625366, |
|
"learning_rate": 7.995222929936306e-05, |
|
"loss": 0.3406, |
|
"step": 1260 |
|
}, |
|
{ |
|
"epoch": 2.04, |
|
"grad_norm": 4.629979133605957, |
|
"learning_rate": 7.963375796178344e-05, |
|
"loss": 0.3548, |
|
"step": 1280 |
|
}, |
|
{ |
|
"epoch": 2.07, |
|
"grad_norm": 0.3911110460758209, |
|
"learning_rate": 7.931528662420382e-05, |
|
"loss": 0.3915, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 2.07, |
|
"eval_accuracy": 0.7638095238095238, |
|
"eval_loss": 0.814107358455658, |
|
"eval_runtime": 15.2728, |
|
"eval_samples_per_second": 68.749, |
|
"eval_steps_per_second": 8.643, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 2.1, |
|
"grad_norm": 12.208237648010254, |
|
"learning_rate": 7.899681528662421e-05, |
|
"loss": 0.3693, |
|
"step": 1320 |
|
}, |
|
{ |
|
"epoch": 2.13, |
|
"grad_norm": 11.460808753967285, |
|
"learning_rate": 7.867834394904459e-05, |
|
"loss": 0.2728, |
|
"step": 1340 |
|
}, |
|
{ |
|
"epoch": 2.17, |
|
"grad_norm": 4.868059158325195, |
|
"learning_rate": 7.835987261146497e-05, |
|
"loss": 0.3227, |
|
"step": 1360 |
|
}, |
|
{ |
|
"epoch": 2.2, |
|
"grad_norm": 2.08876895904541, |
|
"learning_rate": 7.804140127388536e-05, |
|
"loss": 0.2696, |
|
"step": 1380 |
|
}, |
|
{ |
|
"epoch": 2.23, |
|
"grad_norm": 7.153079032897949, |
|
"learning_rate": 7.772292993630573e-05, |
|
"loss": 0.3937, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 2.23, |
|
"eval_accuracy": 0.7704761904761904, |
|
"eval_loss": 0.7881984114646912, |
|
"eval_runtime": 18.2099, |
|
"eval_samples_per_second": 57.661, |
|
"eval_steps_per_second": 7.249, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 2.26, |
|
"grad_norm": 6.202333927154541, |
|
"learning_rate": 7.740445859872611e-05, |
|
"loss": 0.3254, |
|
"step": 1420 |
|
}, |
|
{ |
|
"epoch": 2.29, |
|
"grad_norm": 5.620304107666016, |
|
"learning_rate": 7.70859872611465e-05, |
|
"loss": 0.3754, |
|
"step": 1440 |
|
}, |
|
{ |
|
"epoch": 2.32, |
|
"grad_norm": 2.9809110164642334, |
|
"learning_rate": 7.676751592356688e-05, |
|
"loss": 0.4064, |
|
"step": 1460 |
|
}, |
|
{ |
|
"epoch": 2.36, |
|
"grad_norm": 1.7137058973312378, |
|
"learning_rate": 7.644904458598726e-05, |
|
"loss": 0.3613, |
|
"step": 1480 |
|
}, |
|
{ |
|
"epoch": 2.39, |
|
"grad_norm": 6.261457443237305, |
|
"learning_rate": 7.613057324840765e-05, |
|
"loss": 0.3144, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 2.39, |
|
"eval_accuracy": 0.7514285714285714, |
|
"eval_loss": 0.8657103776931763, |
|
"eval_runtime": 17.3337, |
|
"eval_samples_per_second": 60.576, |
|
"eval_steps_per_second": 7.615, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 2.42, |
|
"grad_norm": 3.399336814880371, |
|
"learning_rate": 7.581210191082803e-05, |
|
"loss": 0.3042, |
|
"step": 1520 |
|
}, |
|
{ |
|
"epoch": 2.45, |
|
"grad_norm": 0.7967382073402405, |
|
"learning_rate": 7.54936305732484e-05, |
|
"loss": 0.291, |
|
"step": 1540 |
|
}, |
|
{ |
|
"epoch": 2.48, |
|
"grad_norm": 6.790742874145508, |
|
"learning_rate": 7.51751592356688e-05, |
|
"loss": 0.3098, |
|
"step": 1560 |
|
}, |
|
{ |
|
"epoch": 2.52, |
|
"grad_norm": 7.324584007263184, |
|
"learning_rate": 7.485668789808917e-05, |
|
"loss": 0.2617, |
|
"step": 1580 |
|
}, |
|
{ |
|
"epoch": 2.55, |
|
"grad_norm": 6.444709300994873, |
|
"learning_rate": 7.453821656050955e-05, |
|
"loss": 0.3143, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 2.55, |
|
"eval_accuracy": 0.7085714285714285, |
|
"eval_loss": 1.056207299232483, |
|
"eval_runtime": 15.1055, |
|
"eval_samples_per_second": 69.511, |
|
"eval_steps_per_second": 8.739, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 2.58, |
|
"grad_norm": 7.054200172424316, |
|
"learning_rate": 7.421974522292993e-05, |
|
"loss": 0.3793, |
|
"step": 1620 |
|
}, |
|
{ |
|
"epoch": 2.61, |
|
"grad_norm": 4.678903579711914, |
|
"learning_rate": 7.390127388535032e-05, |
|
"loss": 0.2708, |
|
"step": 1640 |
|
}, |
|
{ |
|
"epoch": 2.64, |
|
"grad_norm": 10.15630054473877, |
|
"learning_rate": 7.35828025477707e-05, |
|
"loss": 0.3232, |
|
"step": 1660 |
|
}, |
|
{ |
|
"epoch": 2.68, |
|
"grad_norm": 5.110073089599609, |
|
"learning_rate": 7.326433121019108e-05, |
|
"loss": 0.2127, |
|
"step": 1680 |
|
}, |
|
{ |
|
"epoch": 2.71, |
|
"grad_norm": 9.176678657531738, |
|
"learning_rate": 7.294585987261147e-05, |
|
"loss": 0.3884, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 2.71, |
|
"eval_accuracy": 0.7161904761904762, |
|
"eval_loss": 1.0501899719238281, |
|
"eval_runtime": 14.9783, |
|
"eval_samples_per_second": 70.101, |
|
"eval_steps_per_second": 8.813, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 2.74, |
|
"grad_norm": 6.712092876434326, |
|
"learning_rate": 7.262738853503185e-05, |
|
"loss": 0.3172, |
|
"step": 1720 |
|
}, |
|
{ |
|
"epoch": 2.77, |
|
"grad_norm": 4.619245529174805, |
|
"learning_rate": 7.230891719745222e-05, |
|
"loss": 0.3494, |
|
"step": 1740 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"grad_norm": 4.715748310089111, |
|
"learning_rate": 7.199044585987262e-05, |
|
"loss": 0.2987, |
|
"step": 1760 |
|
}, |
|
{ |
|
"epoch": 2.83, |
|
"grad_norm": 0.6376525163650513, |
|
"learning_rate": 7.1671974522293e-05, |
|
"loss": 0.2985, |
|
"step": 1780 |
|
}, |
|
{ |
|
"epoch": 2.87, |
|
"grad_norm": 3.6230549812316895, |
|
"learning_rate": 7.135350318471337e-05, |
|
"loss": 0.3472, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 2.87, |
|
"eval_accuracy": 0.7571428571428571, |
|
"eval_loss": 0.850594699382782, |
|
"eval_runtime": 14.9846, |
|
"eval_samples_per_second": 70.072, |
|
"eval_steps_per_second": 8.809, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 2.9, |
|
"grad_norm": 1.0222142934799194, |
|
"learning_rate": 7.103503184713376e-05, |
|
"loss": 0.2874, |
|
"step": 1820 |
|
}, |
|
{ |
|
"epoch": 2.93, |
|
"grad_norm": 7.561610698699951, |
|
"learning_rate": 7.071656050955414e-05, |
|
"loss": 0.3647, |
|
"step": 1840 |
|
}, |
|
{ |
|
"epoch": 2.96, |
|
"grad_norm": 3.159402370452881, |
|
"learning_rate": 7.039808917197452e-05, |
|
"loss": 0.2743, |
|
"step": 1860 |
|
}, |
|
{ |
|
"epoch": 2.99, |
|
"grad_norm": 1.7933435440063477, |
|
"learning_rate": 7.007961783439491e-05, |
|
"loss": 0.2955, |
|
"step": 1880 |
|
}, |
|
{ |
|
"epoch": 3.03, |
|
"grad_norm": 3.0481631755828857, |
|
"learning_rate": 6.976114649681529e-05, |
|
"loss": 0.2545, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 3.03, |
|
"eval_accuracy": 0.7209523809523809, |
|
"eval_loss": 1.0028949975967407, |
|
"eval_runtime": 14.9856, |
|
"eval_samples_per_second": 70.067, |
|
"eval_steps_per_second": 8.808, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 3.06, |
|
"grad_norm": 2.6126086711883545, |
|
"learning_rate": 6.944267515923567e-05, |
|
"loss": 0.233, |
|
"step": 1920 |
|
}, |
|
{ |
|
"epoch": 3.09, |
|
"grad_norm": 4.197572708129883, |
|
"learning_rate": 6.912420382165606e-05, |
|
"loss": 0.2317, |
|
"step": 1940 |
|
}, |
|
{ |
|
"epoch": 3.12, |
|
"grad_norm": 1.731285810470581, |
|
"learning_rate": 6.880573248407643e-05, |
|
"loss": 0.2237, |
|
"step": 1960 |
|
}, |
|
{ |
|
"epoch": 3.15, |
|
"grad_norm": 2.4551756381988525, |
|
"learning_rate": 6.848726114649681e-05, |
|
"loss": 0.2146, |
|
"step": 1980 |
|
}, |
|
{ |
|
"epoch": 3.18, |
|
"grad_norm": 4.228660583496094, |
|
"learning_rate": 6.81687898089172e-05, |
|
"loss": 0.2213, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 3.18, |
|
"eval_accuracy": 0.7933333333333333, |
|
"eval_loss": 0.8099285960197449, |
|
"eval_runtime": 15.0203, |
|
"eval_samples_per_second": 69.905, |
|
"eval_steps_per_second": 8.788, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 3.22, |
|
"grad_norm": 0.1519007682800293, |
|
"learning_rate": 6.785031847133758e-05, |
|
"loss": 0.2416, |
|
"step": 2020 |
|
}, |
|
{ |
|
"epoch": 3.25, |
|
"grad_norm": 6.512945175170898, |
|
"learning_rate": 6.753184713375796e-05, |
|
"loss": 0.2955, |
|
"step": 2040 |
|
}, |
|
{ |
|
"epoch": 3.28, |
|
"grad_norm": 3.3740453720092773, |
|
"learning_rate": 6.721337579617835e-05, |
|
"loss": 0.2502, |
|
"step": 2060 |
|
}, |
|
{ |
|
"epoch": 3.31, |
|
"grad_norm": 13.082685470581055, |
|
"learning_rate": 6.689490445859873e-05, |
|
"loss": 0.3225, |
|
"step": 2080 |
|
}, |
|
{ |
|
"epoch": 3.34, |
|
"grad_norm": 0.8484336137771606, |
|
"learning_rate": 6.65764331210191e-05, |
|
"loss": 0.3429, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 3.34, |
|
"eval_accuracy": 0.7466666666666667, |
|
"eval_loss": 0.9165627360343933, |
|
"eval_runtime": 15.0876, |
|
"eval_samples_per_second": 69.593, |
|
"eval_steps_per_second": 8.749, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 3.38, |
|
"grad_norm": 2.131553888320923, |
|
"learning_rate": 6.627388535031848e-05, |
|
"loss": 0.2263, |
|
"step": 2120 |
|
}, |
|
{ |
|
"epoch": 3.41, |
|
"grad_norm": 0.06019827350974083, |
|
"learning_rate": 6.595541401273886e-05, |
|
"loss": 0.2397, |
|
"step": 2140 |
|
}, |
|
{ |
|
"epoch": 3.44, |
|
"grad_norm": 6.373557090759277, |
|
"learning_rate": 6.563694267515924e-05, |
|
"loss": 0.2236, |
|
"step": 2160 |
|
}, |
|
{ |
|
"epoch": 3.47, |
|
"grad_norm": 2.4861459732055664, |
|
"learning_rate": 6.531847133757962e-05, |
|
"loss": 0.2515, |
|
"step": 2180 |
|
}, |
|
{ |
|
"epoch": 3.5, |
|
"grad_norm": 11.316793441772461, |
|
"learning_rate": 6.500000000000001e-05, |
|
"loss": 0.3478, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 3.5, |
|
"eval_accuracy": 0.7561904761904762, |
|
"eval_loss": 0.9201664328575134, |
|
"eval_runtime": 15.3295, |
|
"eval_samples_per_second": 68.495, |
|
"eval_steps_per_second": 8.611, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 3.54, |
|
"grad_norm": 6.753748893737793, |
|
"learning_rate": 6.468152866242039e-05, |
|
"loss": 0.2106, |
|
"step": 2220 |
|
}, |
|
{ |
|
"epoch": 3.57, |
|
"grad_norm": 3.0223276615142822, |
|
"learning_rate": 6.436305732484076e-05, |
|
"loss": 0.1678, |
|
"step": 2240 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"grad_norm": 8.31506633758545, |
|
"learning_rate": 6.404458598726115e-05, |
|
"loss": 0.2301, |
|
"step": 2260 |
|
}, |
|
{ |
|
"epoch": 3.63, |
|
"grad_norm": 12.245420455932617, |
|
"learning_rate": 6.372611464968153e-05, |
|
"loss": 0.2038, |
|
"step": 2280 |
|
}, |
|
{ |
|
"epoch": 3.66, |
|
"grad_norm": 4.953712463378906, |
|
"learning_rate": 6.340764331210191e-05, |
|
"loss": 0.2247, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 3.66, |
|
"eval_accuracy": 0.7638095238095238, |
|
"eval_loss": 0.985853910446167, |
|
"eval_runtime": 15.021, |
|
"eval_samples_per_second": 69.902, |
|
"eval_steps_per_second": 8.788, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 3.69, |
|
"grad_norm": 7.013779640197754, |
|
"learning_rate": 6.30891719745223e-05, |
|
"loss": 0.2492, |
|
"step": 2320 |
|
}, |
|
{ |
|
"epoch": 3.73, |
|
"grad_norm": 7.607855319976807, |
|
"learning_rate": 6.277070063694268e-05, |
|
"loss": 0.1915, |
|
"step": 2340 |
|
}, |
|
{ |
|
"epoch": 3.76, |
|
"grad_norm": 7.456921100616455, |
|
"learning_rate": 6.245222929936306e-05, |
|
"loss": 0.2568, |
|
"step": 2360 |
|
}, |
|
{ |
|
"epoch": 3.79, |
|
"grad_norm": 5.894541263580322, |
|
"learning_rate": 6.213375796178345e-05, |
|
"loss": 0.2534, |
|
"step": 2380 |
|
}, |
|
{ |
|
"epoch": 3.82, |
|
"grad_norm": 2.887982130050659, |
|
"learning_rate": 6.181528662420383e-05, |
|
"loss": 0.2873, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 3.82, |
|
"eval_accuracy": 0.7390476190476191, |
|
"eval_loss": 1.0160545110702515, |
|
"eval_runtime": 15.1086, |
|
"eval_samples_per_second": 69.497, |
|
"eval_steps_per_second": 8.737, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 3.85, |
|
"grad_norm": 0.935616672039032, |
|
"learning_rate": 6.14968152866242e-05, |
|
"loss": 0.3611, |
|
"step": 2420 |
|
}, |
|
{ |
|
"epoch": 3.89, |
|
"grad_norm": 1.3130766153335571, |
|
"learning_rate": 6.11783439490446e-05, |
|
"loss": 0.2596, |
|
"step": 2440 |
|
}, |
|
{ |
|
"epoch": 3.92, |
|
"grad_norm": 2.0882582664489746, |
|
"learning_rate": 6.085987261146497e-05, |
|
"loss": 0.311, |
|
"step": 2460 |
|
}, |
|
{ |
|
"epoch": 3.95, |
|
"grad_norm": 5.632941246032715, |
|
"learning_rate": 6.054140127388536e-05, |
|
"loss": 0.2415, |
|
"step": 2480 |
|
}, |
|
{ |
|
"epoch": 3.98, |
|
"grad_norm": 0.2712355852127075, |
|
"learning_rate": 6.0222929936305736e-05, |
|
"loss": 0.2815, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 3.98, |
|
"eval_accuracy": 0.7590476190476191, |
|
"eval_loss": 0.9630508422851562, |
|
"eval_runtime": 15.1527, |
|
"eval_samples_per_second": 69.295, |
|
"eval_steps_per_second": 8.711, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 4.01, |
|
"grad_norm": 1.808802843093872, |
|
"learning_rate": 5.990445859872612e-05, |
|
"loss": 0.2004, |
|
"step": 2520 |
|
}, |
|
{ |
|
"epoch": 4.04, |
|
"grad_norm": 1.8293687105178833, |
|
"learning_rate": 5.9585987261146505e-05, |
|
"loss": 0.2575, |
|
"step": 2540 |
|
}, |
|
{ |
|
"epoch": 4.08, |
|
"grad_norm": 7.782797813415527, |
|
"learning_rate": 5.926751592356688e-05, |
|
"loss": 0.2977, |
|
"step": 2560 |
|
}, |
|
{ |
|
"epoch": 4.11, |
|
"grad_norm": 6.617612838745117, |
|
"learning_rate": 5.894904458598727e-05, |
|
"loss": 0.1844, |
|
"step": 2580 |
|
}, |
|
{ |
|
"epoch": 4.14, |
|
"grad_norm": 2.1140823364257812, |
|
"learning_rate": 5.863057324840765e-05, |
|
"loss": 0.1706, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 4.14, |
|
"eval_accuracy": 0.741904761904762, |
|
"eval_loss": 0.9995649456977844, |
|
"eval_runtime": 14.9483, |
|
"eval_samples_per_second": 70.242, |
|
"eval_steps_per_second": 8.83, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 4.17, |
|
"grad_norm": 3.152398109436035, |
|
"learning_rate": 5.831210191082803e-05, |
|
"loss": 0.2508, |
|
"step": 2620 |
|
}, |
|
{ |
|
"epoch": 4.2, |
|
"grad_norm": 2.778350830078125, |
|
"learning_rate": 5.7993630573248414e-05, |
|
"loss": 0.1961, |
|
"step": 2640 |
|
}, |
|
{ |
|
"epoch": 4.24, |
|
"grad_norm": 2.680100679397583, |
|
"learning_rate": 5.76751592356688e-05, |
|
"loss": 0.2261, |
|
"step": 2660 |
|
}, |
|
{ |
|
"epoch": 4.27, |
|
"grad_norm": 1.4785109758377075, |
|
"learning_rate": 5.7356687898089176e-05, |
|
"loss": 0.3083, |
|
"step": 2680 |
|
}, |
|
{ |
|
"epoch": 4.3, |
|
"grad_norm": 7.928396224975586, |
|
"learning_rate": 5.703821656050956e-05, |
|
"loss": 0.1709, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 4.3, |
|
"eval_accuracy": 0.7028571428571428, |
|
"eval_loss": 1.1969478130340576, |
|
"eval_runtime": 15.008, |
|
"eval_samples_per_second": 69.963, |
|
"eval_steps_per_second": 8.795, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 4.33, |
|
"grad_norm": 4.27495813369751, |
|
"learning_rate": 5.673566878980892e-05, |
|
"loss": 0.2189, |
|
"step": 2720 |
|
}, |
|
{ |
|
"epoch": 4.36, |
|
"grad_norm": 2.0654408931732178, |
|
"learning_rate": 5.6417197452229296e-05, |
|
"loss": 0.1724, |
|
"step": 2740 |
|
}, |
|
{ |
|
"epoch": 4.39, |
|
"grad_norm": 6.015655517578125, |
|
"learning_rate": 5.609872611464968e-05, |
|
"loss": 0.2304, |
|
"step": 2760 |
|
}, |
|
{ |
|
"epoch": 4.43, |
|
"grad_norm": 5.418392658233643, |
|
"learning_rate": 5.5780254777070065e-05, |
|
"loss": 0.2434, |
|
"step": 2780 |
|
}, |
|
{ |
|
"epoch": 4.46, |
|
"grad_norm": 5.330830097198486, |
|
"learning_rate": 5.546178343949044e-05, |
|
"loss": 0.2847, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 4.46, |
|
"eval_accuracy": 0.7276190476190476, |
|
"eval_loss": 1.0895568132400513, |
|
"eval_runtime": 15.0519, |
|
"eval_samples_per_second": 69.759, |
|
"eval_steps_per_second": 8.77, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 4.49, |
|
"grad_norm": 3.495591163635254, |
|
"learning_rate": 5.514331210191083e-05, |
|
"loss": 0.2109, |
|
"step": 2820 |
|
}, |
|
{ |
|
"epoch": 4.52, |
|
"grad_norm": 2.1377482414245605, |
|
"learning_rate": 5.482484076433121e-05, |
|
"loss": 0.2599, |
|
"step": 2840 |
|
}, |
|
{ |
|
"epoch": 4.55, |
|
"grad_norm": 3.6066718101501465, |
|
"learning_rate": 5.450636942675159e-05, |
|
"loss": 0.2052, |
|
"step": 2860 |
|
}, |
|
{ |
|
"epoch": 4.59, |
|
"grad_norm": 8.174120903015137, |
|
"learning_rate": 5.4187898089171974e-05, |
|
"loss": 0.2501, |
|
"step": 2880 |
|
}, |
|
{ |
|
"epoch": 4.62, |
|
"grad_norm": 7.208817481994629, |
|
"learning_rate": 5.386942675159236e-05, |
|
"loss": 0.286, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 4.62, |
|
"eval_accuracy": 0.7628571428571429, |
|
"eval_loss": 0.9894290566444397, |
|
"eval_runtime": 15.1829, |
|
"eval_samples_per_second": 69.157, |
|
"eval_steps_per_second": 8.694, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 4.65, |
|
"grad_norm": 0.47829246520996094, |
|
"learning_rate": 5.3550955414012736e-05, |
|
"loss": 0.202, |
|
"step": 2920 |
|
}, |
|
{ |
|
"epoch": 4.68, |
|
"grad_norm": 4.554569244384766, |
|
"learning_rate": 5.323248407643312e-05, |
|
"loss": 0.225, |
|
"step": 2940 |
|
}, |
|
{ |
|
"epoch": 4.71, |
|
"grad_norm": 7.922834396362305, |
|
"learning_rate": 5.2914012738853506e-05, |
|
"loss": 0.2861, |
|
"step": 2960 |
|
}, |
|
{ |
|
"epoch": 4.75, |
|
"grad_norm": 4.200577735900879, |
|
"learning_rate": 5.2595541401273883e-05, |
|
"loss": 0.173, |
|
"step": 2980 |
|
}, |
|
{ |
|
"epoch": 4.78, |
|
"grad_norm": 6.183908939361572, |
|
"learning_rate": 5.227707006369427e-05, |
|
"loss": 0.2066, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 4.78, |
|
"eval_accuracy": 0.7485714285714286, |
|
"eval_loss": 1.0703792572021484, |
|
"eval_runtime": 15.0438, |
|
"eval_samples_per_second": 69.796, |
|
"eval_steps_per_second": 8.774, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 4.81, |
|
"grad_norm": 8.824482917785645, |
|
"learning_rate": 5.1958598726114646e-05, |
|
"loss": 0.2808, |
|
"step": 3020 |
|
}, |
|
{ |
|
"epoch": 4.84, |
|
"grad_norm": 4.873842716217041, |
|
"learning_rate": 5.164012738853503e-05, |
|
"loss": 0.2838, |
|
"step": 3040 |
|
}, |
|
{ |
|
"epoch": 4.87, |
|
"grad_norm": 4.467446327209473, |
|
"learning_rate": 5.1321656050955415e-05, |
|
"loss": 0.1993, |
|
"step": 3060 |
|
}, |
|
{ |
|
"epoch": 4.9, |
|
"grad_norm": 2.0264174938201904, |
|
"learning_rate": 5.100318471337579e-05, |
|
"loss": 0.1696, |
|
"step": 3080 |
|
}, |
|
{ |
|
"epoch": 4.94, |
|
"grad_norm": 0.3309895396232605, |
|
"learning_rate": 5.068471337579618e-05, |
|
"loss": 0.1579, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 4.94, |
|
"eval_accuracy": 0.780952380952381, |
|
"eval_loss": 0.9727718234062195, |
|
"eval_runtime": 15.0757, |
|
"eval_samples_per_second": 69.648, |
|
"eval_steps_per_second": 8.756, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 4.97, |
|
"grad_norm": 13.229090690612793, |
|
"learning_rate": 5.036624203821656e-05, |
|
"loss": 0.2475, |
|
"step": 3120 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"grad_norm": 0.18536876142024994, |
|
"learning_rate": 5.004777070063694e-05, |
|
"loss": 0.1731, |
|
"step": 3140 |
|
}, |
|
{ |
|
"epoch": 5.03, |
|
"grad_norm": 2.7908613681793213, |
|
"learning_rate": 4.9729299363057324e-05, |
|
"loss": 0.1825, |
|
"step": 3160 |
|
}, |
|
{ |
|
"epoch": 5.06, |
|
"grad_norm": 1.8446813821792603, |
|
"learning_rate": 4.941082802547771e-05, |
|
"loss": 0.225, |
|
"step": 3180 |
|
}, |
|
{ |
|
"epoch": 5.1, |
|
"grad_norm": 1.650841236114502, |
|
"learning_rate": 4.9092356687898087e-05, |
|
"loss": 0.1716, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 5.1, |
|
"eval_accuracy": 0.7123809523809523, |
|
"eval_loss": 1.1833752393722534, |
|
"eval_runtime": 15.2994, |
|
"eval_samples_per_second": 68.63, |
|
"eval_steps_per_second": 8.628, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 5.13, |
|
"grad_norm": 2.1248185634613037, |
|
"learning_rate": 4.877388535031847e-05, |
|
"loss": 0.1548, |
|
"step": 3220 |
|
}, |
|
{ |
|
"epoch": 5.16, |
|
"grad_norm": 1.8586862087249756, |
|
"learning_rate": 4.8455414012738856e-05, |
|
"loss": 0.1676, |
|
"step": 3240 |
|
}, |
|
{ |
|
"epoch": 5.19, |
|
"grad_norm": 0.8100853562355042, |
|
"learning_rate": 4.8136942675159233e-05, |
|
"loss": 0.2012, |
|
"step": 3260 |
|
}, |
|
{ |
|
"epoch": 5.22, |
|
"grad_norm": 4.240858554840088, |
|
"learning_rate": 4.781847133757962e-05, |
|
"loss": 0.2005, |
|
"step": 3280 |
|
}, |
|
{ |
|
"epoch": 5.25, |
|
"grad_norm": 10.562071800231934, |
|
"learning_rate": 4.75e-05, |
|
"loss": 0.1584, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 5.25, |
|
"eval_accuracy": 0.7523809523809524, |
|
"eval_loss": 1.0277926921844482, |
|
"eval_runtime": 15.3062, |
|
"eval_samples_per_second": 68.6, |
|
"eval_steps_per_second": 8.624, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 5.29, |
|
"grad_norm": 0.4224202632904053, |
|
"learning_rate": 4.718152866242038e-05, |
|
"loss": 0.199, |
|
"step": 3320 |
|
}, |
|
{ |
|
"epoch": 5.32, |
|
"grad_norm": 0.20747283101081848, |
|
"learning_rate": 4.6863057324840765e-05, |
|
"loss": 0.2439, |
|
"step": 3340 |
|
}, |
|
{ |
|
"epoch": 5.35, |
|
"grad_norm": 0.2911475896835327, |
|
"learning_rate": 4.654458598726115e-05, |
|
"loss": 0.2808, |
|
"step": 3360 |
|
}, |
|
{ |
|
"epoch": 5.38, |
|
"grad_norm": 4.381645202636719, |
|
"learning_rate": 4.622611464968153e-05, |
|
"loss": 0.1426, |
|
"step": 3380 |
|
}, |
|
{ |
|
"epoch": 5.41, |
|
"grad_norm": 2.27622389793396, |
|
"learning_rate": 4.590764331210191e-05, |
|
"loss": 0.1419, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 5.41, |
|
"eval_accuracy": 0.7371428571428571, |
|
"eval_loss": 1.1342014074325562, |
|
"eval_runtime": 15.0273, |
|
"eval_samples_per_second": 69.873, |
|
"eval_steps_per_second": 8.784, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 5.45, |
|
"grad_norm": 2.8847768306732178, |
|
"learning_rate": 4.5589171974522296e-05, |
|
"loss": 0.2371, |
|
"step": 3420 |
|
}, |
|
{ |
|
"epoch": 5.48, |
|
"grad_norm": 12.182328224182129, |
|
"learning_rate": 4.5270700636942674e-05, |
|
"loss": 0.2688, |
|
"step": 3440 |
|
}, |
|
{ |
|
"epoch": 5.51, |
|
"grad_norm": 5.950127124786377, |
|
"learning_rate": 4.495222929936306e-05, |
|
"loss": 0.1657, |
|
"step": 3460 |
|
}, |
|
{ |
|
"epoch": 5.54, |
|
"grad_norm": 0.8961606621742249, |
|
"learning_rate": 4.4633757961783443e-05, |
|
"loss": 0.1859, |
|
"step": 3480 |
|
}, |
|
{ |
|
"epoch": 5.57, |
|
"grad_norm": 4.591856956481934, |
|
"learning_rate": 4.431528662420382e-05, |
|
"loss": 0.2002, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 5.57, |
|
"eval_accuracy": 0.7323809523809524, |
|
"eval_loss": 1.1519098281860352, |
|
"eval_runtime": 15.3172, |
|
"eval_samples_per_second": 68.551, |
|
"eval_steps_per_second": 8.618, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 5.61, |
|
"grad_norm": 4.491227626800537, |
|
"learning_rate": 4.3996815286624206e-05, |
|
"loss": 0.1545, |
|
"step": 3520 |
|
}, |
|
{ |
|
"epoch": 5.64, |
|
"grad_norm": 10.363092422485352, |
|
"learning_rate": 4.3678343949044584e-05, |
|
"loss": 0.207, |
|
"step": 3540 |
|
}, |
|
{ |
|
"epoch": 5.67, |
|
"grad_norm": 5.781987190246582, |
|
"learning_rate": 4.335987261146497e-05, |
|
"loss": 0.1606, |
|
"step": 3560 |
|
}, |
|
{ |
|
"epoch": 5.7, |
|
"grad_norm": 13.844124794006348, |
|
"learning_rate": 4.304140127388535e-05, |
|
"loss": 0.139, |
|
"step": 3580 |
|
}, |
|
{ |
|
"epoch": 5.73, |
|
"grad_norm": 10.184959411621094, |
|
"learning_rate": 4.272292993630573e-05, |
|
"loss": 0.1987, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 5.73, |
|
"eval_accuracy": 0.7561904761904762, |
|
"eval_loss": 1.0741750001907349, |
|
"eval_runtime": 15.1322, |
|
"eval_samples_per_second": 69.388, |
|
"eval_steps_per_second": 8.723, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 5.76, |
|
"grad_norm": 1.3531529903411865, |
|
"learning_rate": 4.2404458598726115e-05, |
|
"loss": 0.1534, |
|
"step": 3620 |
|
}, |
|
{ |
|
"epoch": 5.8, |
|
"grad_norm": 5.064535617828369, |
|
"learning_rate": 4.20859872611465e-05, |
|
"loss": 0.2151, |
|
"step": 3640 |
|
}, |
|
{ |
|
"epoch": 5.83, |
|
"grad_norm": 3.2219882011413574, |
|
"learning_rate": 4.176751592356688e-05, |
|
"loss": 0.1485, |
|
"step": 3660 |
|
}, |
|
{ |
|
"epoch": 5.86, |
|
"grad_norm": 2.8255577087402344, |
|
"learning_rate": 4.144904458598726e-05, |
|
"loss": 0.1773, |
|
"step": 3680 |
|
}, |
|
{ |
|
"epoch": 5.89, |
|
"grad_norm": 8.52731704711914, |
|
"learning_rate": 4.1130573248407647e-05, |
|
"loss": 0.1207, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 5.89, |
|
"eval_accuracy": 0.7380952380952381, |
|
"eval_loss": 1.1578549146652222, |
|
"eval_runtime": 14.9376, |
|
"eval_samples_per_second": 70.293, |
|
"eval_steps_per_second": 8.837, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 5.92, |
|
"grad_norm": 4.12412166595459, |
|
"learning_rate": 4.0812101910828024e-05, |
|
"loss": 0.1709, |
|
"step": 3720 |
|
}, |
|
{ |
|
"epoch": 5.96, |
|
"grad_norm": 6.599554538726807, |
|
"learning_rate": 4.049363057324841e-05, |
|
"loss": 0.2047, |
|
"step": 3740 |
|
}, |
|
{ |
|
"epoch": 5.99, |
|
"grad_norm": 1.2085940837860107, |
|
"learning_rate": 4.0175159235668793e-05, |
|
"loss": 0.1329, |
|
"step": 3760 |
|
}, |
|
{ |
|
"epoch": 6.02, |
|
"grad_norm": 0.18271246552467346, |
|
"learning_rate": 3.985668789808917e-05, |
|
"loss": 0.3181, |
|
"step": 3780 |
|
}, |
|
{ |
|
"epoch": 6.05, |
|
"grad_norm": 3.431486129760742, |
|
"learning_rate": 3.9538216560509556e-05, |
|
"loss": 0.1403, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 6.05, |
|
"eval_accuracy": 0.7304761904761905, |
|
"eval_loss": 1.100738763809204, |
|
"eval_runtime": 15.064, |
|
"eval_samples_per_second": 69.703, |
|
"eval_steps_per_second": 8.763, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 6.08, |
|
"grad_norm": 2.7748119831085205, |
|
"learning_rate": 3.921974522292994e-05, |
|
"loss": 0.2, |
|
"step": 3820 |
|
}, |
|
{ |
|
"epoch": 6.11, |
|
"grad_norm": 7.583074569702148, |
|
"learning_rate": 3.890127388535032e-05, |
|
"loss": 0.1741, |
|
"step": 3840 |
|
}, |
|
{ |
|
"epoch": 6.15, |
|
"grad_norm": 0.11816076934337616, |
|
"learning_rate": 3.85828025477707e-05, |
|
"loss": 0.1037, |
|
"step": 3860 |
|
}, |
|
{ |
|
"epoch": 6.18, |
|
"grad_norm": 1.9137811660766602, |
|
"learning_rate": 3.826433121019109e-05, |
|
"loss": 0.1443, |
|
"step": 3880 |
|
}, |
|
{ |
|
"epoch": 6.21, |
|
"grad_norm": 0.09991751611232758, |
|
"learning_rate": 3.7945859872611465e-05, |
|
"loss": 0.1569, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 6.21, |
|
"eval_accuracy": 0.7466666666666667, |
|
"eval_loss": 1.112838864326477, |
|
"eval_runtime": 15.0365, |
|
"eval_samples_per_second": 69.83, |
|
"eval_steps_per_second": 8.779, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 6.24, |
|
"grad_norm": 1.1668697595596313, |
|
"learning_rate": 3.762738853503185e-05, |
|
"loss": 0.177, |
|
"step": 3920 |
|
}, |
|
{ |
|
"epoch": 6.27, |
|
"grad_norm": 1.2017813920974731, |
|
"learning_rate": 3.7308917197452234e-05, |
|
"loss": 0.2475, |
|
"step": 3940 |
|
}, |
|
{ |
|
"epoch": 6.31, |
|
"grad_norm": 1.67183256149292, |
|
"learning_rate": 3.699044585987261e-05, |
|
"loss": 0.1248, |
|
"step": 3960 |
|
}, |
|
{ |
|
"epoch": 6.34, |
|
"grad_norm": 3.272127389907837, |
|
"learning_rate": 3.6671974522292997e-05, |
|
"loss": 0.1516, |
|
"step": 3980 |
|
}, |
|
{ |
|
"epoch": 6.37, |
|
"grad_norm": 2.3881561756134033, |
|
"learning_rate": 3.635350318471338e-05, |
|
"loss": 0.1763, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 6.37, |
|
"eval_accuracy": 0.7523809523809524, |
|
"eval_loss": 1.0720144510269165, |
|
"eval_runtime": 15.1243, |
|
"eval_samples_per_second": 69.425, |
|
"eval_steps_per_second": 8.728, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 6.4, |
|
"grad_norm": 3.0372302532196045, |
|
"learning_rate": 3.603503184713376e-05, |
|
"loss": 0.1839, |
|
"step": 4020 |
|
}, |
|
{ |
|
"epoch": 6.43, |
|
"grad_norm": 5.162286281585693, |
|
"learning_rate": 3.5716560509554144e-05, |
|
"loss": 0.2029, |
|
"step": 4040 |
|
}, |
|
{ |
|
"epoch": 6.46, |
|
"grad_norm": 1.8002054691314697, |
|
"learning_rate": 3.539808917197452e-05, |
|
"loss": 0.237, |
|
"step": 4060 |
|
}, |
|
{ |
|
"epoch": 6.5, |
|
"grad_norm": 0.39735618233680725, |
|
"learning_rate": 3.5079617834394906e-05, |
|
"loss": 0.1846, |
|
"step": 4080 |
|
}, |
|
{ |
|
"epoch": 6.53, |
|
"grad_norm": 3.818883180618286, |
|
"learning_rate": 3.476114649681529e-05, |
|
"loss": 0.2426, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 6.53, |
|
"eval_accuracy": 0.7247619047619047, |
|
"eval_loss": 1.1483975648880005, |
|
"eval_runtime": 14.9761, |
|
"eval_samples_per_second": 70.112, |
|
"eval_steps_per_second": 8.814, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 6.56, |
|
"grad_norm": 0.8610227704048157, |
|
"learning_rate": 3.444267515923567e-05, |
|
"loss": 0.1394, |
|
"step": 4120 |
|
}, |
|
{ |
|
"epoch": 6.59, |
|
"grad_norm": 7.635997295379639, |
|
"learning_rate": 3.412420382165605e-05, |
|
"loss": 0.2134, |
|
"step": 4140 |
|
}, |
|
{ |
|
"epoch": 6.62, |
|
"grad_norm": 0.2937486171722412, |
|
"learning_rate": 3.380573248407644e-05, |
|
"loss": 0.1797, |
|
"step": 4160 |
|
}, |
|
{ |
|
"epoch": 6.66, |
|
"grad_norm": 2.422271966934204, |
|
"learning_rate": 3.3487261146496815e-05, |
|
"loss": 0.1803, |
|
"step": 4180 |
|
}, |
|
{ |
|
"epoch": 6.69, |
|
"grad_norm": 6.8127007484436035, |
|
"learning_rate": 3.31687898089172e-05, |
|
"loss": 0.1434, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 6.69, |
|
"eval_accuracy": 0.7342857142857143, |
|
"eval_loss": 1.1789644956588745, |
|
"eval_runtime": 14.9138, |
|
"eval_samples_per_second": 70.405, |
|
"eval_steps_per_second": 8.851, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 6.72, |
|
"grad_norm": 0.9255816340446472, |
|
"learning_rate": 3.2850318471337584e-05, |
|
"loss": 0.2083, |
|
"step": 4220 |
|
}, |
|
{ |
|
"epoch": 6.75, |
|
"grad_norm": 1.347322702407837, |
|
"learning_rate": 3.253184713375796e-05, |
|
"loss": 0.2296, |
|
"step": 4240 |
|
}, |
|
{ |
|
"epoch": 6.78, |
|
"grad_norm": 0.05040664225816727, |
|
"learning_rate": 3.221337579617835e-05, |
|
"loss": 0.1151, |
|
"step": 4260 |
|
}, |
|
{ |
|
"epoch": 6.82, |
|
"grad_norm": 0.04670969396829605, |
|
"learning_rate": 3.189490445859873e-05, |
|
"loss": 0.1435, |
|
"step": 4280 |
|
}, |
|
{ |
|
"epoch": 6.85, |
|
"grad_norm": 3.0971717834472656, |
|
"learning_rate": 3.157643312101911e-05, |
|
"loss": 0.2191, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 6.85, |
|
"eval_accuracy": 0.7485714285714286, |
|
"eval_loss": 1.1169472932815552, |
|
"eval_runtime": 14.9321, |
|
"eval_samples_per_second": 70.319, |
|
"eval_steps_per_second": 8.84, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 6.88, |
|
"grad_norm": 0.6114574670791626, |
|
"learning_rate": 3.1257961783439494e-05, |
|
"loss": 0.1114, |
|
"step": 4320 |
|
}, |
|
{ |
|
"epoch": 6.91, |
|
"grad_norm": 4.660861015319824, |
|
"learning_rate": 3.093949044585988e-05, |
|
"loss": 0.186, |
|
"step": 4340 |
|
}, |
|
{ |
|
"epoch": 6.94, |
|
"grad_norm": 1.0122532844543457, |
|
"learning_rate": 3.0621019108280256e-05, |
|
"loss": 0.1284, |
|
"step": 4360 |
|
}, |
|
{ |
|
"epoch": 6.97, |
|
"grad_norm": 4.9644880294799805, |
|
"learning_rate": 3.030254777070064e-05, |
|
"loss": 0.2162, |
|
"step": 4380 |
|
}, |
|
{ |
|
"epoch": 7.01, |
|
"grad_norm": 3.802830219268799, |
|
"learning_rate": 2.998407643312102e-05, |
|
"loss": 0.2062, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 7.01, |
|
"eval_accuracy": 0.7609523809523809, |
|
"eval_loss": 1.1300264596939087, |
|
"eval_runtime": 14.9402, |
|
"eval_samples_per_second": 70.28, |
|
"eval_steps_per_second": 8.835, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 7.04, |
|
"grad_norm": 0.5785458087921143, |
|
"learning_rate": 2.9665605095541403e-05, |
|
"loss": 0.082, |
|
"step": 4420 |
|
}, |
|
{ |
|
"epoch": 7.07, |
|
"grad_norm": 0.0573742650449276, |
|
"learning_rate": 2.9347133757961787e-05, |
|
"loss": 0.1013, |
|
"step": 4440 |
|
}, |
|
{ |
|
"epoch": 7.1, |
|
"grad_norm": 2.6736879348754883, |
|
"learning_rate": 2.902866242038217e-05, |
|
"loss": 0.1665, |
|
"step": 4460 |
|
}, |
|
{ |
|
"epoch": 7.13, |
|
"grad_norm": 4.041416645050049, |
|
"learning_rate": 2.871019108280255e-05, |
|
"loss": 0.1626, |
|
"step": 4480 |
|
}, |
|
{ |
|
"epoch": 7.17, |
|
"grad_norm": 1.28642737865448, |
|
"learning_rate": 2.8391719745222934e-05, |
|
"loss": 0.1495, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 7.17, |
|
"eval_accuracy": 0.7495238095238095, |
|
"eval_loss": 1.1477110385894775, |
|
"eval_runtime": 14.8067, |
|
"eval_samples_per_second": 70.914, |
|
"eval_steps_per_second": 8.915, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 7.2, |
|
"grad_norm": 2.69707989692688, |
|
"learning_rate": 2.8073248407643316e-05, |
|
"loss": 0.1423, |
|
"step": 4520 |
|
}, |
|
{ |
|
"epoch": 7.23, |
|
"grad_norm": 6.730169296264648, |
|
"learning_rate": 2.7754777070063697e-05, |
|
"loss": 0.1623, |
|
"step": 4540 |
|
}, |
|
{ |
|
"epoch": 7.26, |
|
"grad_norm": 7.955442428588867, |
|
"learning_rate": 2.743630573248408e-05, |
|
"loss": 0.2045, |
|
"step": 4560 |
|
}, |
|
{ |
|
"epoch": 7.29, |
|
"grad_norm": 3.3925881385803223, |
|
"learning_rate": 2.7117834394904462e-05, |
|
"loss": 0.0884, |
|
"step": 4580 |
|
}, |
|
{ |
|
"epoch": 7.32, |
|
"grad_norm": 12.87939167022705, |
|
"learning_rate": 2.6799363057324844e-05, |
|
"loss": 0.1261, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 7.32, |
|
"eval_accuracy": 0.7657142857142857, |
|
"eval_loss": 1.0891320705413818, |
|
"eval_runtime": 15.1965, |
|
"eval_samples_per_second": 69.095, |
|
"eval_steps_per_second": 8.686, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 7.36, |
|
"grad_norm": 1.0216339826583862, |
|
"learning_rate": 2.6480891719745228e-05, |
|
"loss": 0.1678, |
|
"step": 4620 |
|
}, |
|
{ |
|
"epoch": 7.39, |
|
"grad_norm": 3.8394949436187744, |
|
"learning_rate": 2.616242038216561e-05, |
|
"loss": 0.129, |
|
"step": 4640 |
|
}, |
|
{ |
|
"epoch": 7.42, |
|
"grad_norm": 2.563333511352539, |
|
"learning_rate": 2.584394904458599e-05, |
|
"loss": 0.1512, |
|
"step": 4660 |
|
}, |
|
{ |
|
"epoch": 7.45, |
|
"grad_norm": 4.966637134552002, |
|
"learning_rate": 2.5525477707006372e-05, |
|
"loss": 0.1261, |
|
"step": 4680 |
|
}, |
|
{ |
|
"epoch": 7.48, |
|
"grad_norm": 6.724792003631592, |
|
"learning_rate": 2.5207006369426756e-05, |
|
"loss": 0.12, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 7.48, |
|
"eval_accuracy": 0.76, |
|
"eval_loss": 1.1359137296676636, |
|
"eval_runtime": 14.87, |
|
"eval_samples_per_second": 70.612, |
|
"eval_steps_per_second": 8.877, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 7.52, |
|
"grad_norm": 0.4127102196216583, |
|
"learning_rate": 2.4888535031847134e-05, |
|
"loss": 0.1206, |
|
"step": 4720 |
|
}, |
|
{ |
|
"epoch": 7.55, |
|
"grad_norm": 0.21055227518081665, |
|
"learning_rate": 2.457006369426752e-05, |
|
"loss": 0.1455, |
|
"step": 4740 |
|
}, |
|
{ |
|
"epoch": 7.58, |
|
"grad_norm": 3.864630937576294, |
|
"learning_rate": 2.426751592356688e-05, |
|
"loss": 0.1809, |
|
"step": 4760 |
|
}, |
|
{ |
|
"epoch": 7.61, |
|
"grad_norm": 2.548668146133423, |
|
"learning_rate": 2.3949044585987263e-05, |
|
"loss": 0.182, |
|
"step": 4780 |
|
}, |
|
{ |
|
"epoch": 7.64, |
|
"grad_norm": 5.192627906799316, |
|
"learning_rate": 2.3630573248407645e-05, |
|
"loss": 0.1396, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 7.64, |
|
"eval_accuracy": 0.7409523809523809, |
|
"eval_loss": 1.1230192184448242, |
|
"eval_runtime": 15.1532, |
|
"eval_samples_per_second": 69.292, |
|
"eval_steps_per_second": 8.711, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 7.68, |
|
"grad_norm": 4.093466281890869, |
|
"learning_rate": 2.3312101910828026e-05, |
|
"loss": 0.1563, |
|
"step": 4820 |
|
}, |
|
{ |
|
"epoch": 7.71, |
|
"grad_norm": 0.3261258006095886, |
|
"learning_rate": 2.299363057324841e-05, |
|
"loss": 0.1461, |
|
"step": 4840 |
|
}, |
|
{ |
|
"epoch": 7.74, |
|
"grad_norm": 1.178528070449829, |
|
"learning_rate": 2.267515923566879e-05, |
|
"loss": 0.1578, |
|
"step": 4860 |
|
}, |
|
{ |
|
"epoch": 7.77, |
|
"grad_norm": 1.2750300168991089, |
|
"learning_rate": 2.2356687898089173e-05, |
|
"loss": 0.1937, |
|
"step": 4880 |
|
}, |
|
{ |
|
"epoch": 7.8, |
|
"grad_norm": 5.664385795593262, |
|
"learning_rate": 2.2038216560509557e-05, |
|
"loss": 0.0728, |
|
"step": 4900 |
|
}, |
|
{ |
|
"epoch": 7.8, |
|
"eval_accuracy": 0.7552380952380953, |
|
"eval_loss": 1.120953917503357, |
|
"eval_runtime": 15.3497, |
|
"eval_samples_per_second": 68.405, |
|
"eval_steps_per_second": 8.6, |
|
"step": 4900 |
|
}, |
|
{ |
|
"epoch": 7.83, |
|
"grad_norm": 0.04029368981719017, |
|
"learning_rate": 2.171974522292994e-05, |
|
"loss": 0.0901, |
|
"step": 4920 |
|
}, |
|
{ |
|
"epoch": 7.87, |
|
"grad_norm": 0.18388265371322632, |
|
"learning_rate": 2.140127388535032e-05, |
|
"loss": 0.0982, |
|
"step": 4940 |
|
}, |
|
{ |
|
"epoch": 7.9, |
|
"grad_norm": 0.03308388963341713, |
|
"learning_rate": 2.10828025477707e-05, |
|
"loss": 0.1396, |
|
"step": 4960 |
|
}, |
|
{ |
|
"epoch": 7.93, |
|
"grad_norm": 0.023461850360035896, |
|
"learning_rate": 2.0764331210191085e-05, |
|
"loss": 0.2196, |
|
"step": 4980 |
|
}, |
|
{ |
|
"epoch": 7.96, |
|
"grad_norm": 0.02594674378633499, |
|
"learning_rate": 2.0445859872611467e-05, |
|
"loss": 0.175, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 7.96, |
|
"eval_accuracy": 0.7504761904761905, |
|
"eval_loss": 1.1203620433807373, |
|
"eval_runtime": 14.9947, |
|
"eval_samples_per_second": 70.025, |
|
"eval_steps_per_second": 8.803, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 7.99, |
|
"grad_norm": 0.27780893445014954, |
|
"learning_rate": 2.0127388535031848e-05, |
|
"loss": 0.1099, |
|
"step": 5020 |
|
}, |
|
{ |
|
"epoch": 8.03, |
|
"grad_norm": 0.02279558964073658, |
|
"learning_rate": 1.9808917197452232e-05, |
|
"loss": 0.1265, |
|
"step": 5040 |
|
}, |
|
{ |
|
"epoch": 8.06, |
|
"grad_norm": 4.824516773223877, |
|
"learning_rate": 1.9490445859872614e-05, |
|
"loss": 0.0849, |
|
"step": 5060 |
|
}, |
|
{ |
|
"epoch": 8.09, |
|
"grad_norm": 5.1949334144592285, |
|
"learning_rate": 1.9171974522292995e-05, |
|
"loss": 0.1568, |
|
"step": 5080 |
|
}, |
|
{ |
|
"epoch": 8.12, |
|
"grad_norm": 1.2651612758636475, |
|
"learning_rate": 1.885350318471338e-05, |
|
"loss": 0.1214, |
|
"step": 5100 |
|
}, |
|
{ |
|
"epoch": 8.12, |
|
"eval_accuracy": 0.7542857142857143, |
|
"eval_loss": 1.106448769569397, |
|
"eval_runtime": 14.9134, |
|
"eval_samples_per_second": 70.407, |
|
"eval_steps_per_second": 8.851, |
|
"step": 5100 |
|
}, |
|
{ |
|
"epoch": 8.15, |
|
"grad_norm": 3.7672173976898193, |
|
"learning_rate": 1.8535031847133757e-05, |
|
"loss": 0.1851, |
|
"step": 5120 |
|
}, |
|
{ |
|
"epoch": 8.18, |
|
"grad_norm": 0.10589015483856201, |
|
"learning_rate": 1.8216560509554138e-05, |
|
"loss": 0.0865, |
|
"step": 5140 |
|
}, |
|
{ |
|
"epoch": 8.22, |
|
"grad_norm": 3.8695759773254395, |
|
"learning_rate": 1.7898089171974523e-05, |
|
"loss": 0.1378, |
|
"step": 5160 |
|
}, |
|
{ |
|
"epoch": 8.25, |
|
"grad_norm": 0.10915953665971756, |
|
"learning_rate": 1.7579617834394904e-05, |
|
"loss": 0.108, |
|
"step": 5180 |
|
}, |
|
{ |
|
"epoch": 8.28, |
|
"grad_norm": 6.51039981842041, |
|
"learning_rate": 1.7261146496815285e-05, |
|
"loss": 0.1218, |
|
"step": 5200 |
|
}, |
|
{ |
|
"epoch": 8.28, |
|
"eval_accuracy": 0.7771428571428571, |
|
"eval_loss": 1.023201823234558, |
|
"eval_runtime": 14.7617, |
|
"eval_samples_per_second": 71.13, |
|
"eval_steps_per_second": 8.942, |
|
"step": 5200 |
|
}, |
|
{ |
|
"epoch": 8.31, |
|
"grad_norm": 5.766481876373291, |
|
"learning_rate": 1.694267515923567e-05, |
|
"loss": 0.1055, |
|
"step": 5220 |
|
}, |
|
{ |
|
"epoch": 8.34, |
|
"grad_norm": 1.3557641506195068, |
|
"learning_rate": 1.662420382165605e-05, |
|
"loss": 0.1426, |
|
"step": 5240 |
|
}, |
|
{ |
|
"epoch": 8.38, |
|
"grad_norm": 2.0178778171539307, |
|
"learning_rate": 1.6305732484076432e-05, |
|
"loss": 0.1147, |
|
"step": 5260 |
|
}, |
|
{ |
|
"epoch": 8.41, |
|
"grad_norm": 0.3355507254600525, |
|
"learning_rate": 1.5987261146496817e-05, |
|
"loss": 0.0986, |
|
"step": 5280 |
|
}, |
|
{ |
|
"epoch": 8.44, |
|
"grad_norm": 8.486790657043457, |
|
"learning_rate": 1.5668789808917198e-05, |
|
"loss": 0.1556, |
|
"step": 5300 |
|
}, |
|
{ |
|
"epoch": 8.44, |
|
"eval_accuracy": 0.7771428571428571, |
|
"eval_loss": 1.0489068031311035, |
|
"eval_runtime": 14.8821, |
|
"eval_samples_per_second": 70.554, |
|
"eval_steps_per_second": 8.87, |
|
"step": 5300 |
|
}, |
|
{ |
|
"epoch": 8.47, |
|
"grad_norm": 0.0707675963640213, |
|
"learning_rate": 1.536624203821656e-05, |
|
"loss": 0.0841, |
|
"step": 5320 |
|
}, |
|
{ |
|
"epoch": 8.5, |
|
"grad_norm": 16.97945213317871, |
|
"learning_rate": 1.5047770700636943e-05, |
|
"loss": 0.1825, |
|
"step": 5340 |
|
}, |
|
{ |
|
"epoch": 8.54, |
|
"grad_norm": 3.0360000133514404, |
|
"learning_rate": 1.4729299363057326e-05, |
|
"loss": 0.1569, |
|
"step": 5360 |
|
}, |
|
{ |
|
"epoch": 8.57, |
|
"grad_norm": 0.4013698697090149, |
|
"learning_rate": 1.4410828025477707e-05, |
|
"loss": 0.115, |
|
"step": 5380 |
|
}, |
|
{ |
|
"epoch": 8.6, |
|
"grad_norm": 0.027194073423743248, |
|
"learning_rate": 1.409235668789809e-05, |
|
"loss": 0.1019, |
|
"step": 5400 |
|
}, |
|
{ |
|
"epoch": 8.6, |
|
"eval_accuracy": 0.7752380952380953, |
|
"eval_loss": 1.091567873954773, |
|
"eval_runtime": 15.0217, |
|
"eval_samples_per_second": 69.899, |
|
"eval_steps_per_second": 8.787, |
|
"step": 5400 |
|
}, |
|
{ |
|
"epoch": 8.63, |
|
"grad_norm": 5.168393611907959, |
|
"learning_rate": 1.3773885350318472e-05, |
|
"loss": 0.0843, |
|
"step": 5420 |
|
}, |
|
{ |
|
"epoch": 8.66, |
|
"grad_norm": 0.056524887681007385, |
|
"learning_rate": 1.3455414012738854e-05, |
|
"loss": 0.0594, |
|
"step": 5440 |
|
}, |
|
{ |
|
"epoch": 8.69, |
|
"grad_norm": 0.16345758736133575, |
|
"learning_rate": 1.3136942675159237e-05, |
|
"loss": 0.1234, |
|
"step": 5460 |
|
}, |
|
{ |
|
"epoch": 8.73, |
|
"grad_norm": 9.12987995147705, |
|
"learning_rate": 1.2818471337579618e-05, |
|
"loss": 0.0643, |
|
"step": 5480 |
|
}, |
|
{ |
|
"epoch": 8.76, |
|
"grad_norm": 9.363956451416016, |
|
"learning_rate": 1.25e-05, |
|
"loss": 0.1446, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 8.76, |
|
"eval_accuracy": 0.7504761904761905, |
|
"eval_loss": 1.1855816841125488, |
|
"eval_runtime": 15.0734, |
|
"eval_samples_per_second": 69.659, |
|
"eval_steps_per_second": 8.757, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 8.79, |
|
"grad_norm": 0.7227168083190918, |
|
"learning_rate": 1.2181528662420383e-05, |
|
"loss": 0.1234, |
|
"step": 5520 |
|
}, |
|
{ |
|
"epoch": 8.82, |
|
"grad_norm": 2.766003370285034, |
|
"learning_rate": 1.1863057324840765e-05, |
|
"loss": 0.0921, |
|
"step": 5540 |
|
}, |
|
{ |
|
"epoch": 8.85, |
|
"grad_norm": 0.011736826971173286, |
|
"learning_rate": 1.1544585987261148e-05, |
|
"loss": 0.0684, |
|
"step": 5560 |
|
}, |
|
{ |
|
"epoch": 8.89, |
|
"grad_norm": 0.02954920195043087, |
|
"learning_rate": 1.1226114649681529e-05, |
|
"loss": 0.109, |
|
"step": 5580 |
|
}, |
|
{ |
|
"epoch": 8.92, |
|
"grad_norm": 1.0454398393630981, |
|
"learning_rate": 1.0907643312101912e-05, |
|
"loss": 0.1348, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 8.92, |
|
"eval_accuracy": 0.7638095238095238, |
|
"eval_loss": 1.1379646062850952, |
|
"eval_runtime": 14.986, |
|
"eval_samples_per_second": 70.065, |
|
"eval_steps_per_second": 8.808, |
|
"step": 5600 |
|
}, |
|
{ |
|
"epoch": 8.95, |
|
"grad_norm": 0.1124209389090538, |
|
"learning_rate": 1.0589171974522294e-05, |
|
"loss": 0.1118, |
|
"step": 5620 |
|
}, |
|
{ |
|
"epoch": 8.98, |
|
"grad_norm": 0.05434571951627731, |
|
"learning_rate": 1.0270700636942676e-05, |
|
"loss": 0.0981, |
|
"step": 5640 |
|
}, |
|
{ |
|
"epoch": 9.01, |
|
"grad_norm": 6.469759941101074, |
|
"learning_rate": 9.952229299363059e-06, |
|
"loss": 0.1224, |
|
"step": 5660 |
|
}, |
|
{ |
|
"epoch": 9.04, |
|
"grad_norm": 2.5620501041412354, |
|
"learning_rate": 9.633757961783441e-06, |
|
"loss": 0.0646, |
|
"step": 5680 |
|
}, |
|
{ |
|
"epoch": 9.08, |
|
"grad_norm": 4.989463806152344, |
|
"learning_rate": 9.315286624203821e-06, |
|
"loss": 0.1402, |
|
"step": 5700 |
|
}, |
|
{ |
|
"epoch": 9.08, |
|
"eval_accuracy": 0.7695238095238095, |
|
"eval_loss": 1.1233410835266113, |
|
"eval_runtime": 15.0736, |
|
"eval_samples_per_second": 69.658, |
|
"eval_steps_per_second": 8.757, |
|
"step": 5700 |
|
}, |
|
{ |
|
"epoch": 9.11, |
|
"grad_norm": 0.28552043437957764, |
|
"learning_rate": 8.996815286624204e-06, |
|
"loss": 0.0686, |
|
"step": 5720 |
|
}, |
|
{ |
|
"epoch": 9.14, |
|
"grad_norm": 0.09686419367790222, |
|
"learning_rate": 8.678343949044587e-06, |
|
"loss": 0.1217, |
|
"step": 5740 |
|
}, |
|
{ |
|
"epoch": 9.17, |
|
"grad_norm": 5.7156572341918945, |
|
"learning_rate": 8.359872611464968e-06, |
|
"loss": 0.1461, |
|
"step": 5760 |
|
}, |
|
{ |
|
"epoch": 9.2, |
|
"grad_norm": 10.184377670288086, |
|
"learning_rate": 8.04140127388535e-06, |
|
"loss": 0.0892, |
|
"step": 5780 |
|
}, |
|
{ |
|
"epoch": 9.24, |
|
"grad_norm": 0.3750424087047577, |
|
"learning_rate": 7.722929936305732e-06, |
|
"loss": 0.1075, |
|
"step": 5800 |
|
}, |
|
{ |
|
"epoch": 9.24, |
|
"eval_accuracy": 0.7628571428571429, |
|
"eval_loss": 1.1471986770629883, |
|
"eval_runtime": 15.0092, |
|
"eval_samples_per_second": 69.957, |
|
"eval_steps_per_second": 8.795, |
|
"step": 5800 |
|
}, |
|
{ |
|
"epoch": 9.27, |
|
"grad_norm": 12.310994148254395, |
|
"learning_rate": 7.404458598726115e-06, |
|
"loss": 0.1211, |
|
"step": 5820 |
|
}, |
|
{ |
|
"epoch": 9.3, |
|
"grad_norm": 0.08023884147405624, |
|
"learning_rate": 7.085987261146497e-06, |
|
"loss": 0.0503, |
|
"step": 5840 |
|
}, |
|
{ |
|
"epoch": 9.33, |
|
"grad_norm": 10.179062843322754, |
|
"learning_rate": 6.76751592356688e-06, |
|
"loss": 0.0965, |
|
"step": 5860 |
|
}, |
|
{ |
|
"epoch": 9.36, |
|
"grad_norm": 6.414700031280518, |
|
"learning_rate": 6.449044585987262e-06, |
|
"loss": 0.1485, |
|
"step": 5880 |
|
}, |
|
{ |
|
"epoch": 9.39, |
|
"grad_norm": 4.7796854972839355, |
|
"learning_rate": 6.130573248407644e-06, |
|
"loss": 0.0991, |
|
"step": 5900 |
|
}, |
|
{ |
|
"epoch": 9.39, |
|
"eval_accuracy": 0.7647619047619048, |
|
"eval_loss": 1.1529643535614014, |
|
"eval_runtime": 14.9351, |
|
"eval_samples_per_second": 70.304, |
|
"eval_steps_per_second": 8.838, |
|
"step": 5900 |
|
}, |
|
{ |
|
"epoch": 9.43, |
|
"grad_norm": 3.3885576725006104, |
|
"learning_rate": 5.812101910828026e-06, |
|
"loss": 0.0582, |
|
"step": 5920 |
|
}, |
|
{ |
|
"epoch": 9.46, |
|
"grad_norm": 0.3516397178173065, |
|
"learning_rate": 5.493630573248408e-06, |
|
"loss": 0.1151, |
|
"step": 5940 |
|
}, |
|
{ |
|
"epoch": 9.49, |
|
"grad_norm": 0.029496684670448303, |
|
"learning_rate": 5.175159235668791e-06, |
|
"loss": 0.0601, |
|
"step": 5960 |
|
}, |
|
{ |
|
"epoch": 9.52, |
|
"grad_norm": 0.00963157694786787, |
|
"learning_rate": 4.856687898089173e-06, |
|
"loss": 0.0885, |
|
"step": 5980 |
|
}, |
|
{ |
|
"epoch": 9.55, |
|
"grad_norm": 1.4221217632293701, |
|
"learning_rate": 4.538216560509554e-06, |
|
"loss": 0.081, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 9.55, |
|
"eval_accuracy": 0.7628571428571429, |
|
"eval_loss": 1.1586228609085083, |
|
"eval_runtime": 14.9129, |
|
"eval_samples_per_second": 70.409, |
|
"eval_steps_per_second": 8.851, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 9.59, |
|
"grad_norm": 0.09046982228755951, |
|
"learning_rate": 4.219745222929937e-06, |
|
"loss": 0.0744, |
|
"step": 6020 |
|
}, |
|
{ |
|
"epoch": 9.62, |
|
"grad_norm": 3.131237030029297, |
|
"learning_rate": 3.901273885350319e-06, |
|
"loss": 0.095, |
|
"step": 6040 |
|
}, |
|
{ |
|
"epoch": 9.65, |
|
"grad_norm": 0.04757837578654289, |
|
"learning_rate": 3.5828025477707007e-06, |
|
"loss": 0.0827, |
|
"step": 6060 |
|
}, |
|
{ |
|
"epoch": 9.68, |
|
"grad_norm": 0.19970394670963287, |
|
"learning_rate": 3.2643312101910827e-06, |
|
"loss": 0.1016, |
|
"step": 6080 |
|
}, |
|
{ |
|
"epoch": 9.71, |
|
"grad_norm": 4.088714599609375, |
|
"learning_rate": 2.945859872611465e-06, |
|
"loss": 0.0724, |
|
"step": 6100 |
|
}, |
|
{ |
|
"epoch": 9.71, |
|
"eval_accuracy": 0.7676190476190476, |
|
"eval_loss": 1.1590627431869507, |
|
"eval_runtime": 14.9656, |
|
"eval_samples_per_second": 70.161, |
|
"eval_steps_per_second": 8.82, |
|
"step": 6100 |
|
}, |
|
{ |
|
"epoch": 9.75, |
|
"grad_norm": 0.02539316564798355, |
|
"learning_rate": 2.6273885350318472e-06, |
|
"loss": 0.0873, |
|
"step": 6120 |
|
}, |
|
{ |
|
"epoch": 9.78, |
|
"grad_norm": 0.04443074390292168, |
|
"learning_rate": 2.3089171974522293e-06, |
|
"loss": 0.1642, |
|
"step": 6140 |
|
}, |
|
{ |
|
"epoch": 9.81, |
|
"grad_norm": 4.564943790435791, |
|
"learning_rate": 1.9904458598726113e-06, |
|
"loss": 0.0766, |
|
"step": 6160 |
|
}, |
|
{ |
|
"epoch": 9.84, |
|
"grad_norm": 3.6170880794525146, |
|
"learning_rate": 1.6719745222929937e-06, |
|
"loss": 0.129, |
|
"step": 6180 |
|
}, |
|
{ |
|
"epoch": 9.87, |
|
"grad_norm": 3.5646533966064453, |
|
"learning_rate": 1.353503184713376e-06, |
|
"loss": 0.0399, |
|
"step": 6200 |
|
}, |
|
{ |
|
"epoch": 9.87, |
|
"eval_accuracy": 0.7695238095238095, |
|
"eval_loss": 1.1417580842971802, |
|
"eval_runtime": 14.9815, |
|
"eval_samples_per_second": 70.086, |
|
"eval_steps_per_second": 8.811, |
|
"step": 6200 |
|
}, |
|
{ |
|
"epoch": 9.9, |
|
"grad_norm": 0.011990565806627274, |
|
"learning_rate": 1.035031847133758e-06, |
|
"loss": 0.079, |
|
"step": 6220 |
|
}, |
|
{ |
|
"epoch": 9.94, |
|
"grad_norm": 0.08819901943206787, |
|
"learning_rate": 7.165605095541401e-07, |
|
"loss": 0.1412, |
|
"step": 6240 |
|
}, |
|
{ |
|
"epoch": 9.97, |
|
"grad_norm": 1.3580721616744995, |
|
"learning_rate": 3.980891719745223e-07, |
|
"loss": 0.0805, |
|
"step": 6260 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 0.38720598816871643, |
|
"learning_rate": 7.961783439490447e-08, |
|
"loss": 0.0744, |
|
"step": 6280 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"step": 6280, |
|
"total_flos": 7.776878731479245e+18, |
|
"train_loss": 0.29035600812002355, |
|
"train_runtime": 3700.3159, |
|
"train_samples_per_second": 27.119, |
|
"train_steps_per_second": 1.697 |
|
} |
|
], |
|
"logging_steps": 20, |
|
"max_steps": 6280, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 10, |
|
"save_steps": 100, |
|
"total_flos": 7.776878731479245e+18, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|