| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.321127579192096, | |
| "eval_steps": 1000000, | |
| "global_step": 64148, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07265329845975008, | |
| "grad_norm": 0.39395979046821594, | |
| "learning_rate": 9.992734670154026e-06, | |
| "loss": 8.1621, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.14530659691950015, | |
| "grad_norm": 0.5780870318412781, | |
| "learning_rate": 9.985469340308051e-06, | |
| "loss": 7.0196, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.21795989537925023, | |
| "grad_norm": 0.5540204048156738, | |
| "learning_rate": 9.978204010462076e-06, | |
| "loss": 6.6142, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2906131938390003, | |
| "grad_norm": 0.5443109273910522, | |
| "learning_rate": 9.970938680616102e-06, | |
| "loss": 6.3813, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.36326649229875035, | |
| "grad_norm": 0.6196579337120056, | |
| "learning_rate": 9.963673350770125e-06, | |
| "loss": 6.2207, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.43591979075850046, | |
| "grad_norm": 0.5789604187011719, | |
| "learning_rate": 9.95640802092415e-06, | |
| "loss": 6.0957, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5085730892182505, | |
| "grad_norm": 0.63157719373703, | |
| "learning_rate": 9.949142691078175e-06, | |
| "loss": 5.9962, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5812263876780006, | |
| "grad_norm": 0.5619252920150757, | |
| "learning_rate": 9.9418773612322e-06, | |
| "loss": 5.9116, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.6538796861377506, | |
| "grad_norm": 0.6025997996330261, | |
| "learning_rate": 9.934612031386226e-06, | |
| "loss": 5.8332, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.7265329845975007, | |
| "grad_norm": 0.6860383152961731, | |
| "learning_rate": 9.927346701540251e-06, | |
| "loss": 5.7695, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.7991862830572508, | |
| "grad_norm": 0.7735297679901123, | |
| "learning_rate": 9.920081371694276e-06, | |
| "loss": 5.7068, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.8718395815170009, | |
| "grad_norm": 0.6047775745391846, | |
| "learning_rate": 9.912816041848301e-06, | |
| "loss": 5.65, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.9444928799767509, | |
| "grad_norm": 0.8214231133460999, | |
| "learning_rate": 9.905550712002325e-06, | |
| "loss": 5.6052, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.017146178436501, | |
| "grad_norm": 0.8017473816871643, | |
| "learning_rate": 9.89828538215635e-06, | |
| "loss": 5.5515, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.0897994768962511, | |
| "grad_norm": 0.7803598642349243, | |
| "learning_rate": 9.891020052310377e-06, | |
| "loss": 5.5058, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.1624527753560012, | |
| "grad_norm": 0.7102193236351013, | |
| "learning_rate": 9.8837547224644e-06, | |
| "loss": 5.4684, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.2351060738157513, | |
| "grad_norm": 0.7121894359588623, | |
| "learning_rate": 9.876489392618425e-06, | |
| "loss": 5.4271, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.3077593722755014, | |
| "grad_norm": 0.7547861337661743, | |
| "learning_rate": 9.86922406277245e-06, | |
| "loss": 5.397, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.3804126707352513, | |
| "grad_norm": 0.7697557210922241, | |
| "learning_rate": 9.861958732926476e-06, | |
| "loss": 5.3563, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.4530659691950014, | |
| "grad_norm": 0.8195068836212158, | |
| "learning_rate": 9.854693403080501e-06, | |
| "loss": 5.3228, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.5257192676547515, | |
| "grad_norm": 0.7637478113174438, | |
| "learning_rate": 9.847428073234524e-06, | |
| "loss": 5.2946, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.5983725661145016, | |
| "grad_norm": 0.8972837328910828, | |
| "learning_rate": 9.840162743388551e-06, | |
| "loss": 5.2588, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.6710258645742515, | |
| "grad_norm": 0.9061026573181152, | |
| "learning_rate": 9.832897413542576e-06, | |
| "loss": 5.2321, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.7436791630340016, | |
| "grad_norm": 0.7589295506477356, | |
| "learning_rate": 9.8256320836966e-06, | |
| "loss": 5.2065, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.8163324614937517, | |
| "grad_norm": 0.8363276720046997, | |
| "learning_rate": 9.818366753850625e-06, | |
| "loss": 5.1819, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.8889857599535018, | |
| "grad_norm": 0.8336675763130188, | |
| "learning_rate": 9.81110142400465e-06, | |
| "loss": 5.1575, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.961639058413252, | |
| "grad_norm": 0.8027577996253967, | |
| "learning_rate": 9.803836094158675e-06, | |
| "loss": 5.1314, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.034292356873002, | |
| "grad_norm": 0.7707503437995911, | |
| "learning_rate": 9.7965707643127e-06, | |
| "loss": 5.1088, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.106945655332752, | |
| "grad_norm": 0.8998398184776306, | |
| "learning_rate": 9.789305434466726e-06, | |
| "loss": 5.0808, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.1795989537925022, | |
| "grad_norm": 0.8320822715759277, | |
| "learning_rate": 9.782040104620751e-06, | |
| "loss": 5.0634, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.2522522522522523, | |
| "grad_norm": 0.8281899094581604, | |
| "learning_rate": 9.774774774774776e-06, | |
| "loss": 5.0444, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.3249055507120024, | |
| "grad_norm": 0.8026254177093506, | |
| "learning_rate": 9.767509444928801e-06, | |
| "loss": 5.0179, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.3975588491717525, | |
| "grad_norm": 0.776054322719574, | |
| "learning_rate": 9.760244115082825e-06, | |
| "loss": 5.0009, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.4702121476315027, | |
| "grad_norm": 0.8994656801223755, | |
| "learning_rate": 9.75297878523685e-06, | |
| "loss": 4.9812, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.5428654460912528, | |
| "grad_norm": 0.8583637475967407, | |
| "learning_rate": 9.745713455390875e-06, | |
| "loss": 4.9628, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.615518744551003, | |
| "grad_norm": 0.893622636795044, | |
| "learning_rate": 9.7384481255449e-06, | |
| "loss": 4.9464, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.688172043010753, | |
| "grad_norm": 0.8372634053230286, | |
| "learning_rate": 9.731182795698925e-06, | |
| "loss": 4.9304, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.7608253414705026, | |
| "grad_norm": 0.9035200476646423, | |
| "learning_rate": 9.72391746585295e-06, | |
| "loss": 4.9127, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.8334786399302527, | |
| "grad_norm": 0.8660902380943298, | |
| "learning_rate": 9.716652136006976e-06, | |
| "loss": 4.8998, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.906131938390003, | |
| "grad_norm": 0.9463688731193542, | |
| "learning_rate": 9.709386806161001e-06, | |
| "loss": 4.8778, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.978785236849753, | |
| "grad_norm": 0.8671707510948181, | |
| "learning_rate": 9.702121476315024e-06, | |
| "loss": 4.8658, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.051438535309503, | |
| "grad_norm": 0.9653003215789795, | |
| "learning_rate": 9.69485614646905e-06, | |
| "loss": 4.853, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.124091833769253, | |
| "grad_norm": 0.8984787464141846, | |
| "learning_rate": 9.687590816623077e-06, | |
| "loss": 4.8364, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.1967451322290033, | |
| "grad_norm": 0.8803947567939758, | |
| "learning_rate": 9.6803254867771e-06, | |
| "loss": 4.8177, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.2693984306887534, | |
| "grad_norm": 0.8390781283378601, | |
| "learning_rate": 9.673060156931125e-06, | |
| "loss": 4.8077, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 3.3420517291485035, | |
| "grad_norm": 0.9399817585945129, | |
| "learning_rate": 9.66579482708515e-06, | |
| "loss": 4.7942, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 3.4147050276082536, | |
| "grad_norm": 0.864647626876831, | |
| "learning_rate": 9.658529497239176e-06, | |
| "loss": 4.7819, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 3.4873583260680037, | |
| "grad_norm": 0.863230288028717, | |
| "learning_rate": 9.6512641673932e-06, | |
| "loss": 4.7675, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 3.5600116245277533, | |
| "grad_norm": 0.9053711891174316, | |
| "learning_rate": 9.643998837547224e-06, | |
| "loss": 4.7562, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 3.6326649229875034, | |
| "grad_norm": 0.9143590331077576, | |
| "learning_rate": 9.636733507701251e-06, | |
| "loss": 4.7408, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 3.7053182214472535, | |
| "grad_norm": 0.8819567561149597, | |
| "learning_rate": 9.629468177855276e-06, | |
| "loss": 4.7347, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 3.7779715199070036, | |
| "grad_norm": 0.9433513283729553, | |
| "learning_rate": 9.6222028480093e-06, | |
| "loss": 4.7227, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 3.8506248183667537, | |
| "grad_norm": 0.9881005883216858, | |
| "learning_rate": 9.614937518163325e-06, | |
| "loss": 4.7096, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 3.923278116826504, | |
| "grad_norm": 0.890139639377594, | |
| "learning_rate": 9.60767218831735e-06, | |
| "loss": 4.7, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 3.995931415286254, | |
| "grad_norm": 0.9619746804237366, | |
| "learning_rate": 9.600406858471375e-06, | |
| "loss": 4.6879, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 4.068584713746004, | |
| "grad_norm": 0.8994712233543396, | |
| "learning_rate": 9.5931415286254e-06, | |
| "loss": 4.6762, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 4.141238012205754, | |
| "grad_norm": 0.9636611938476562, | |
| "learning_rate": 9.585876198779426e-06, | |
| "loss": 4.6643, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 4.213891310665504, | |
| "grad_norm": 0.8346700072288513, | |
| "learning_rate": 9.57861086893345e-06, | |
| "loss": 4.6541, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 4.286544609125254, | |
| "grad_norm": 0.8618379831314087, | |
| "learning_rate": 9.571345539087476e-06, | |
| "loss": 4.6473, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 4.3591979075850045, | |
| "grad_norm": 0.9965615272521973, | |
| "learning_rate": 9.564080209241501e-06, | |
| "loss": 4.636, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 4.431851206044755, | |
| "grad_norm": 0.9595836400985718, | |
| "learning_rate": 9.556814879395525e-06, | |
| "loss": 4.6262, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 4.504504504504505, | |
| "grad_norm": 0.918498694896698, | |
| "learning_rate": 9.54954954954955e-06, | |
| "loss": 4.6139, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 4.577157802964255, | |
| "grad_norm": 0.9323708415031433, | |
| "learning_rate": 9.542284219703575e-06, | |
| "loss": 4.6138, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 4.649811101424005, | |
| "grad_norm": 0.9415541291236877, | |
| "learning_rate": 9.5350188898576e-06, | |
| "loss": 4.6008, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 4.722464399883755, | |
| "grad_norm": 0.9424421787261963, | |
| "learning_rate": 9.527753560011625e-06, | |
| "loss": 4.5893, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 4.795117698343505, | |
| "grad_norm": 0.942669689655304, | |
| "learning_rate": 9.52048823016565e-06, | |
| "loss": 4.5866, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 4.867770996803255, | |
| "grad_norm": 0.9807618856430054, | |
| "learning_rate": 9.513222900319676e-06, | |
| "loss": 4.5736, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 4.940424295263005, | |
| "grad_norm": 0.960895299911499, | |
| "learning_rate": 9.5059575704737e-06, | |
| "loss": 4.5619, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 5.013077593722755, | |
| "grad_norm": 0.9801125526428223, | |
| "learning_rate": 9.498692240627724e-06, | |
| "loss": 4.5551, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 5.0857308921825055, | |
| "grad_norm": 0.9244375228881836, | |
| "learning_rate": 9.49142691078175e-06, | |
| "loss": 4.5408, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 5.158384190642255, | |
| "grad_norm": 0.9349271059036255, | |
| "learning_rate": 9.484161580935776e-06, | |
| "loss": 4.5358, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 5.231037489102005, | |
| "grad_norm": 0.9496700763702393, | |
| "learning_rate": 9.4768962510898e-06, | |
| "loss": 4.5284, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 5.303690787561755, | |
| "grad_norm": 0.8824469447135925, | |
| "learning_rate": 9.469630921243825e-06, | |
| "loss": 4.5225, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 5.376344086021505, | |
| "grad_norm": 0.9746178984642029, | |
| "learning_rate": 9.46236559139785e-06, | |
| "loss": 4.5084, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 5.448997384481255, | |
| "grad_norm": 0.9742151498794556, | |
| "learning_rate": 9.455100261551875e-06, | |
| "loss": 4.5036, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 5.521650682941005, | |
| "grad_norm": 1.0122525691986084, | |
| "learning_rate": 9.4478349317059e-06, | |
| "loss": 4.4951, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 5.594303981400755, | |
| "grad_norm": 0.9454242587089539, | |
| "learning_rate": 9.440569601859924e-06, | |
| "loss": 4.4844, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 5.6669572798605055, | |
| "grad_norm": 0.9594370722770691, | |
| "learning_rate": 9.43330427201395e-06, | |
| "loss": 4.4776, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 5.739610578320256, | |
| "grad_norm": 0.9644564986228943, | |
| "learning_rate": 9.426038942167976e-06, | |
| "loss": 4.4748, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 5.812263876780006, | |
| "grad_norm": 0.9564418792724609, | |
| "learning_rate": 9.418773612322e-06, | |
| "loss": 4.4592, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 5.884917175239756, | |
| "grad_norm": 0.9269986152648926, | |
| "learning_rate": 9.411508282476025e-06, | |
| "loss": 4.455, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 5.957570473699506, | |
| "grad_norm": 0.9712278842926025, | |
| "learning_rate": 9.40424295263005e-06, | |
| "loss": 4.4455, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 6.030223772159256, | |
| "grad_norm": 0.9372689723968506, | |
| "learning_rate": 9.396977622784075e-06, | |
| "loss": 4.4318, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 6.102877070619006, | |
| "grad_norm": 0.942115843296051, | |
| "learning_rate": 9.3897122929381e-06, | |
| "loss": 4.4264, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 6.175530369078756, | |
| "grad_norm": 1.0541362762451172, | |
| "learning_rate": 9.382446963092125e-06, | |
| "loss": 4.4163, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 6.248183667538506, | |
| "grad_norm": 0.9615015387535095, | |
| "learning_rate": 9.37518163324615e-06, | |
| "loss": 4.4096, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 6.320836965998256, | |
| "grad_norm": 1.0179786682128906, | |
| "learning_rate": 9.367916303400176e-06, | |
| "loss": 4.4034, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 6.3934902644580065, | |
| "grad_norm": 0.9816853404045105, | |
| "learning_rate": 9.360650973554201e-06, | |
| "loss": 4.3956, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 6.466143562917757, | |
| "grad_norm": 1.0108842849731445, | |
| "learning_rate": 9.353385643708224e-06, | |
| "loss": 4.3811, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 6.538796861377507, | |
| "grad_norm": 1.0273536443710327, | |
| "learning_rate": 9.34612031386225e-06, | |
| "loss": 4.3748, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 6.611450159837257, | |
| "grad_norm": 0.9984000325202942, | |
| "learning_rate": 9.338854984016275e-06, | |
| "loss": 4.3711, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 6.684103458297007, | |
| "grad_norm": 0.9320149421691895, | |
| "learning_rate": 9.3315896541703e-06, | |
| "loss": 4.3621, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 6.756756756756757, | |
| "grad_norm": 1.1042180061340332, | |
| "learning_rate": 9.324324324324325e-06, | |
| "loss": 4.3516, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 6.829410055216507, | |
| "grad_norm": 0.9890114665031433, | |
| "learning_rate": 9.31705899447835e-06, | |
| "loss": 4.3464, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 6.902063353676257, | |
| "grad_norm": 1.034157156944275, | |
| "learning_rate": 9.309793664632375e-06, | |
| "loss": 4.3374, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 6.974716652136007, | |
| "grad_norm": 0.9817051887512207, | |
| "learning_rate": 9.3025283347864e-06, | |
| "loss": 4.3284, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 7.047369950595757, | |
| "grad_norm": 1.0558934211730957, | |
| "learning_rate": 9.295263004940424e-06, | |
| "loss": 4.3154, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 7.1200232490555075, | |
| "grad_norm": 0.9428159594535828, | |
| "learning_rate": 9.28799767509445e-06, | |
| "loss": 4.3088, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 7.192676547515257, | |
| "grad_norm": 1.0073195695877075, | |
| "learning_rate": 9.280732345248476e-06, | |
| "loss": 4.2983, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 7.265329845975007, | |
| "grad_norm": 0.9252042174339294, | |
| "learning_rate": 9.2734670154025e-06, | |
| "loss": 4.2912, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 7.337983144434757, | |
| "grad_norm": 1.0058454275131226, | |
| "learning_rate": 9.266201685556525e-06, | |
| "loss": 4.2814, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 7.410636442894507, | |
| "grad_norm": 1.0393654108047485, | |
| "learning_rate": 9.25893635571055e-06, | |
| "loss": 4.2693, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 7.483289741354257, | |
| "grad_norm": 0.9429093599319458, | |
| "learning_rate": 9.251671025864575e-06, | |
| "loss": 4.2604, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 7.555943039814007, | |
| "grad_norm": 0.9522086381912231, | |
| "learning_rate": 9.2444056960186e-06, | |
| "loss": 4.2499, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 7.628596338273757, | |
| "grad_norm": 0.9854516386985779, | |
| "learning_rate": 9.237140366172624e-06, | |
| "loss": 4.2411, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 7.7012496367335075, | |
| "grad_norm": 0.9321109056472778, | |
| "learning_rate": 9.22987503632665e-06, | |
| "loss": 4.2257, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 7.773902935193258, | |
| "grad_norm": 0.9426267147064209, | |
| "learning_rate": 9.222609706480676e-06, | |
| "loss": 4.2225, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 7.846556233653008, | |
| "grad_norm": 0.9784730672836304, | |
| "learning_rate": 9.2153443766347e-06, | |
| "loss": 4.2077, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 7.919209532112758, | |
| "grad_norm": 0.9224970936775208, | |
| "learning_rate": 9.208079046788724e-06, | |
| "loss": 4.2029, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 7.991862830572508, | |
| "grad_norm": 0.9461036920547485, | |
| "learning_rate": 9.20081371694275e-06, | |
| "loss": 4.1932, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 8.064516129032258, | |
| "grad_norm": 0.9398289322853088, | |
| "learning_rate": 9.193548387096775e-06, | |
| "loss": 4.182, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 8.137169427492008, | |
| "grad_norm": 0.9409332275390625, | |
| "learning_rate": 9.1862830572508e-06, | |
| "loss": 4.1696, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 8.209822725951758, | |
| "grad_norm": 0.9343535304069519, | |
| "learning_rate": 9.179017727404825e-06, | |
| "loss": 4.1627, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 8.282476024411508, | |
| "grad_norm": 0.96622234582901, | |
| "learning_rate": 9.17175239755885e-06, | |
| "loss": 4.1569, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 8.355129322871258, | |
| "grad_norm": 0.9742193818092346, | |
| "learning_rate": 9.164487067712876e-06, | |
| "loss": 4.1494, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 8.427782621331009, | |
| "grad_norm": 0.9126195907592773, | |
| "learning_rate": 9.1572217378669e-06, | |
| "loss": 4.1413, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 8.500435919790759, | |
| "grad_norm": 0.978380024433136, | |
| "learning_rate": 9.149956408020924e-06, | |
| "loss": 4.1355, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 8.573089218250509, | |
| "grad_norm": 0.943013608455658, | |
| "learning_rate": 9.14269107817495e-06, | |
| "loss": 4.1277, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 8.645742516710259, | |
| "grad_norm": 0.9417549967765808, | |
| "learning_rate": 9.135425748328975e-06, | |
| "loss": 4.1254, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 8.718395815170009, | |
| "grad_norm": 0.9427275061607361, | |
| "learning_rate": 9.128160418483e-06, | |
| "loss": 4.1183, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 8.791049113629759, | |
| "grad_norm": 0.972618818283081, | |
| "learning_rate": 9.120895088637025e-06, | |
| "loss": 4.1094, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 8.86370241208951, | |
| "grad_norm": 0.9143289923667908, | |
| "learning_rate": 9.11362975879105e-06, | |
| "loss": 4.1035, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 8.93635571054926, | |
| "grad_norm": 0.961862325668335, | |
| "learning_rate": 9.106364428945075e-06, | |
| "loss": 4.1035, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 9.00900900900901, | |
| "grad_norm": 0.9252375960350037, | |
| "learning_rate": 9.0990990990991e-06, | |
| "loss": 4.0923, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 9.08166230746876, | |
| "grad_norm": 0.9321468472480774, | |
| "learning_rate": 9.091833769253124e-06, | |
| "loss": 4.0855, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 9.15431560592851, | |
| "grad_norm": 0.9290264844894409, | |
| "learning_rate": 9.084568439407149e-06, | |
| "loss": 4.0847, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 9.22696890438826, | |
| "grad_norm": 1.0008461475372314, | |
| "learning_rate": 9.077303109561176e-06, | |
| "loss": 4.0791, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 9.29962220284801, | |
| "grad_norm": 1.0144729614257812, | |
| "learning_rate": 9.0700377797152e-06, | |
| "loss": 4.0724, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 9.321127579192096, | |
| "step": 64148, | |
| "total_flos": 5.363892569191219e+17, | |
| "train_loss": 4.763968430426513, | |
| "train_runtime": 93599.1484, | |
| "train_samples_per_second": 705.822, | |
| "train_steps_per_second": 7.353 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 688200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.363892569191219e+17, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |