| { | |
| "best_metric": 8.547043800354004, | |
| "best_model_checkpoint": "tiny_bert_rand_10_v1/checkpoint-10000", | |
| "epoch": 25.0, | |
| "eval_steps": 10000, | |
| "global_step": 59550, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.20990764063811923, | |
| "grad_norm": 2.4543728828430176, | |
| "learning_rate": 5e-06, | |
| "loss": 12.0526, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 1.7795683145523071, | |
| "learning_rate": 1e-05, | |
| "loss": 10.5001, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 1.5096403360366821, | |
| "learning_rate": 1.5e-05, | |
| "loss": 9.5134, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 1.6482595205307007, | |
| "learning_rate": 2e-05, | |
| "loss": 9.3153, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 1.3180246353149414, | |
| "learning_rate": 2.5e-05, | |
| "loss": 9.2001, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 1.9059840440750122, | |
| "learning_rate": 3e-05, | |
| "loss": 9.0969, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 1.9212170839309692, | |
| "learning_rate": 3.5e-05, | |
| "loss": 9.0101, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 1.763610601425171, | |
| "learning_rate": 4e-05, | |
| "loss": 8.9272, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 1.122094750404358, | |
| "learning_rate": 4.5e-05, | |
| "loss": 8.8518, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 1.5978599786758423, | |
| "learning_rate": 5e-05, | |
| "loss": 8.7884, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 1.7132242918014526, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 8.7419, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 2.3705999851226807, | |
| "learning_rate": 6e-05, | |
| "loss": 8.7049, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 1.458178162574768, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 8.6729, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 1.7663618326187134, | |
| "learning_rate": 7e-05, | |
| "loss": 8.6442, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 1.9466015100479126, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 8.6011, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 1.9804456233978271, | |
| "learning_rate": 8e-05, | |
| "loss": 8.5708, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 2.040253162384033, | |
| "learning_rate": 8.5e-05, | |
| "loss": 8.5569, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 1.698929786682129, | |
| "learning_rate": 9e-05, | |
| "loss": 8.5328, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 2.173619031906128, | |
| "learning_rate": 9.5e-05, | |
| "loss": 8.5175, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 2.3768362998962402, | |
| "learning_rate": 0.0001, | |
| "loss": 8.4297, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15061632210133838, | |
| "eval_loss": 8.547043800354004, | |
| "eval_runtime": 0.7286, | |
| "eval_samples_per_second": 657.465, | |
| "eval_steps_per_second": 6.863, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 2.380319833755493, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 8.4291, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 2.2213191986083984, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 8.4233, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 2.3587307929992676, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 8.4191, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 2.3701372146606445, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 8.3966, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 2.6558446884155273, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 8.3142, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 2.040811538696289, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 8.3218, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 2.4577434062957764, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 8.3266, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 1.7516306638717651, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 8.3294, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 2.5566599369049072, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 8.2847, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 2.2998855113983154, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 8.2482, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 2.8845982551574707, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 8.2473, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 2.732081174850464, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 8.2581, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 2.6092946529388428, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 8.2574, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 2.636404275894165, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 8.2045, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 2.5597710609436035, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 8.181, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 2.1486318111419678, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 8.1943, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 2.652672290802002, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 8.2057, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 2.148256301879883, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 8.2056, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 2.4641363620758057, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 8.1354, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 2.7930519580841064, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 8.1384, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.15266982799741854, | |
| "eval_loss": 8.619112014770508, | |
| "eval_runtime": 0.7044, | |
| "eval_samples_per_second": 680.006, | |
| "eval_steps_per_second": 7.098, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 2.4523539543151855, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 8.1414, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 2.3080801963806152, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 8.1552, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 2.7334702014923096, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 8.1381, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 2.830592393875122, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 8.0742, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 2.6776628494262695, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 8.0806, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 2.213458776473999, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 8.0933, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 3.096095323562622, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 8.0891, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 2.957338809967041, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 8.0564, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 3.1426753997802734, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 8.0005, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 2.555790424346924, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 8.0028, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 2.698042154312134, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 8.0104, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 2.4399921894073486, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 8.0229, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 3.3951821327209473, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 7.9427, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 3.1927943229675293, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 7.8975, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 4.027744770050049, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 7.9167, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 3.2401740550994873, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 7.9207, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 2.755052328109741, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 7.9376, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 3.7951042652130127, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 7.7978, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 4.011223316192627, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 7.784, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 3.801516532897949, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 7.8049, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15328650977978006, | |
| "eval_loss": 8.925741195678711, | |
| "eval_runtime": 0.7018, | |
| "eval_samples_per_second": 682.56, | |
| "eval_steps_per_second": 7.125, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 3.4474925994873047, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 7.8213, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 4.206075191497803, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 7.8314, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 4.515190124511719, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 7.6447, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 5.009239196777344, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 7.6711, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 4.6359782218933105, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 7.7005, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 4.072754383087158, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 7.7116, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 5.391729831695557, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 7.6545, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 5.038398265838623, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 7.5266, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 5.284417629241943, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 7.5438, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 4.765394687652588, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 7.5684, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 5.750059127807617, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 7.585, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 5.908089637756348, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 7.4718, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 4.953010559082031, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 7.4035, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 4.632134914398193, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 7.429, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 5.068370819091797, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 7.4396, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 5.3794426918029785, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 7.4649, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 6.545327663421631, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 7.3019, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 6.5143632888793945, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 7.2861, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 6.060126781463623, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 7.3126, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 5.459360599517822, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 7.317, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15190086351912077, | |
| "eval_loss": 9.591103553771973, | |
| "eval_runtime": 0.7023, | |
| "eval_samples_per_second": 682.085, | |
| "eval_steps_per_second": 7.12, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 6.614737510681152, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 7.3382, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 7.430929660797119, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 7.1516, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 6.5889739990234375, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 7.1671, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 7.773727893829346, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 7.1901, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 7.578391075134277, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 7.2116, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 6.625092506408691, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 7.1665, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 6.528718948364258, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 7.0465, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 8.716383934020996, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 7.0621, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 6.42665433883667, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 7.0906, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 6.703744888305664, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 7.1147, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 6.843780517578125, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 7.0281, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 7.743956565856934, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 6.9609, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 8.763398170471191, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 6.9692, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 6.573347568511963, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 6.9814, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 7.759286403656006, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 6.9936, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 8.392936706542969, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 6.9053, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 9.589981079101562, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 6.868, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 8.077651023864746, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 6.8873, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 8.214678764343262, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 6.8996, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 9.277266502380371, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 6.9107, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1510359716264334, | |
| "eval_loss": 10.22226619720459, | |
| "eval_runtime": 0.7089, | |
| "eval_samples_per_second": 675.722, | |
| "eval_steps_per_second": 7.053, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 8.224175453186035, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 6.7989, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 8.156539916992188, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 6.8082, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 8.12301254272461, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 6.8176, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 8.275008201599121, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 6.8233, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 8.482283592224121, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 6.8118, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 8.714301109313965, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 6.7316, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 8.544315338134766, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 6.7433, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 8.602791786193848, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 6.7491, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 7.862239360809326, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 6.7496, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 7.541048526763916, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 6.728, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 7.04449987411499, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 6.6838, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 8.786259651184082, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 6.6968, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 7.5606560707092285, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 6.6892, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 9.471149444580078, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 6.6989, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 7.306352138519287, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 6.6666, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 7.249099254608154, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 6.6675, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 9.488431930541992, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 6.659, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 8.312141418457031, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 6.6544, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 9.254082679748535, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 6.6449, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 3.050247306064896e+17, | |
| "train_loss": 7.781117795084225, | |
| "train_runtime": 11819.9466, | |
| "train_samples_per_second": 483.587, | |
| "train_steps_per_second": 5.038 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "total_flos": 3.050247306064896e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |