| { | |
| "best_metric": 0.011125968769192696, | |
| "best_model_checkpoint": "./mistral-magyar-portas-results/checkpoint-2500", | |
| "epoch": 2.7429467084639496, | |
| "eval_steps": 500, | |
| "global_step": 3500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03918495297805643, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0137, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07836990595611286, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0186, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11755485893416928, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 0.00019991124482238458, | |
| "loss": 0.0345, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15673981191222572, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 0.00019964513683916945, | |
| "loss": 0.0232, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19592476489028213, | |
| "grad_norm": 0.09228515625, | |
| "learning_rate": 0.00019920214841958082, | |
| "loss": 0.0165, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.23510971786833856, | |
| "grad_norm": 0.0537109375, | |
| "learning_rate": 0.00019858306591393602, | |
| "loss": 0.0145, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.274294670846395, | |
| "grad_norm": 0.09130859375, | |
| "learning_rate": 0.00019778898825778996, | |
| "loss": 0.0136, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.31347962382445144, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 0.00019682132502121086, | |
| "loss": 0.0135, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.35266457680250785, | |
| "grad_norm": 0.2373046875, | |
| "learning_rate": 0.00019568179390664744, | |
| "loss": 0.0169, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.39184952978056425, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 0.00019437241769982907, | |
| "loss": 0.0323, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.39184952978056425, | |
| "eval_loss": 0.06335489451885223, | |
| "eval_runtime": 438.733, | |
| "eval_samples_per_second": 2.587, | |
| "eval_steps_per_second": 1.295, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.43103448275862066, | |
| "grad_norm": 0.115234375, | |
| "learning_rate": 0.00019289552067911186, | |
| "loss": 0.0273, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4702194357366771, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 0.00019125372448964363, | |
| "loss": 0.0197, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5094043887147336, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 0.00018944994348967247, | |
| "loss": 0.0201, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.54858934169279, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 0.00018748737957725904, | |
| "loss": 0.022, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5877742946708464, | |
| "grad_norm": 0.0888671875, | |
| "learning_rate": 0.00018536951650657585, | |
| "loss": 0.0157, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6269592476489029, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 0.00018310011370388304, | |
| "loss": 0.0144, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6661442006269592, | |
| "grad_norm": 0.08984375, | |
| "learning_rate": 0.00018068319959415723, | |
| "loss": 0.0149, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7053291536050157, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 0.00017812306445022025, | |
| "loss": 0.0137, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7445141065830722, | |
| "grad_norm": 0.05810546875, | |
| "learning_rate": 0.0001754242527770605, | |
| "loss": 0.0142, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7836990595611285, | |
| "grad_norm": 0.10791015625, | |
| "learning_rate": 0.000172591555244866, | |
| "loss": 0.0123, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7836990595611285, | |
| "eval_loss": 0.013054505921900272, | |
| "eval_runtime": 438.9352, | |
| "eval_samples_per_second": 2.586, | |
| "eval_steps_per_second": 1.294, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.822884012539185, | |
| "grad_norm": 0.08447265625, | |
| "learning_rate": 0.0001696300001850887, | |
| "loss": 0.013, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 0.0615234375, | |
| "learning_rate": 0.0001665448446646357, | |
| "loss": 0.0127, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9012539184952978, | |
| "grad_norm": 0.08642578125, | |
| "learning_rate": 0.00016334156515403065, | |
| "loss": 0.0124, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9404388714733543, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 0.00016002584780611194, | |
| "loss": 0.0124, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9796238244514106, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 0.00015660357836252232, | |
| "loss": 0.0127, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0188087774294672, | |
| "grad_norm": 0.10400390625, | |
| "learning_rate": 0.000153080831705908, | |
| "loss": 0.0107, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.0579937304075235, | |
| "grad_norm": 0.099609375, | |
| "learning_rate": 0.00014946386107637306, | |
| "loss": 0.0092, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.09717868338558, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 0.00014575908697133058, | |
| "loss": 0.0093, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 0.00014197308574845488, | |
| "loss": 0.0099, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.1755485893416928, | |
| "grad_norm": 0.0869140625, | |
| "learning_rate": 0.0001381125779519658, | |
| "loss": 0.0094, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.1755485893416928, | |
| "eval_loss": 0.012080053798854351, | |
| "eval_runtime": 438.8603, | |
| "eval_samples_per_second": 2.586, | |
| "eval_steps_per_second": 1.294, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.2147335423197492, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 0.00013418441638296652, | |
| "loss": 0.0095, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.2539184952978055, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 0.00013019557393501228, | |
| "loss": 0.0093, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.293103448275862, | |
| "grad_norm": 0.0537109375, | |
| "learning_rate": 0.00012615313121650204, | |
| "loss": 0.0098, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.3322884012539185, | |
| "grad_norm": 0.057861328125, | |
| "learning_rate": 0.00012206426398186534, | |
| "loss": 0.01, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.3714733542319748, | |
| "grad_norm": 0.0693359375, | |
| "learning_rate": 0.00011793623039385545, | |
| "loss": 0.0094, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.4106583072100314, | |
| "grad_norm": 0.12109375, | |
| "learning_rate": 0.00011377635813955834, | |
| "loss": 0.0096, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.4498432601880877, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 0.00010959203142298981, | |
| "loss": 0.0091, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.489028213166144, | |
| "grad_norm": 0.050537109375, | |
| "learning_rate": 0.00010539067785736856, | |
| "loss": 0.0086, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.5282131661442007, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 0.0001011797552803333, | |
| "loss": 0.0091, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.567398119122257, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 9.696673851550907e-05, | |
| "loss": 0.0093, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.567398119122257, | |
| "eval_loss": 0.011984162032604218, | |
| "eval_runtime": 438.9114, | |
| "eval_samples_per_second": 2.586, | |
| "eval_steps_per_second": 1.294, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6065830721003134, | |
| "grad_norm": 0.09814453125, | |
| "learning_rate": 9.275910610392104e-05, | |
| "loss": 0.0091, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.64576802507837, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 8.856432702880984e-05, | |
| "loss": 0.0096, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.6849529780564263, | |
| "grad_norm": 0.07568359375, | |
| "learning_rate": 8.43898474574128e-05, | |
| "loss": 0.0092, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 8.02430775232462e-05, | |
| "loss": 0.0084, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.7633228840125392, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 7.61313781723508e-05, | |
| "loss": 0.0087, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.8025078369905956, | |
| "grad_norm": 0.06591796875, | |
| "learning_rate": 7.206204809685029e-05, | |
| "loss": 0.0086, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.841692789968652, | |
| "grad_norm": 0.052978515625, | |
| "learning_rate": 6.804231077901733e-05, | |
| "loss": 0.0085, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.8808777429467085, | |
| "grad_norm": 0.07421875, | |
| "learning_rate": 6.407930166884409e-05, | |
| "loss": 0.0094, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.9200626959247649, | |
| "grad_norm": 0.1044921875, | |
| "learning_rate": 6.018005551787984e-05, | |
| "loss": 0.008, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.9592476489028212, | |
| "grad_norm": 0.05615234375, | |
| "learning_rate": 5.635149389181855e-05, | |
| "loss": 0.0088, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.9592476489028212, | |
| "eval_loss": 0.011125968769192696, | |
| "eval_runtime": 438.9494, | |
| "eval_samples_per_second": 2.586, | |
| "eval_steps_per_second": 1.294, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.9984326018808778, | |
| "grad_norm": 0.038330078125, | |
| "learning_rate": 5.260041288400284e-05, | |
| "loss": 0.0092, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.0376175548589344, | |
| "grad_norm": 0.056396484375, | |
| "learning_rate": 4.893347105165468e-05, | |
| "loss": 0.0064, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.0768025078369905, | |
| "grad_norm": 0.0517578125, | |
| "learning_rate": 4.535717759624677e-05, | |
| "loss": 0.0065, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.115987460815047, | |
| "grad_norm": 0.049560546875, | |
| "learning_rate": 4.187788080899591e-05, | |
| "loss": 0.0064, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.1551724137931036, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 3.8501756801988675e-05, | |
| "loss": 0.0062, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.19435736677116, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 3.5234798544942914e-05, | |
| "loss": 0.0063, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.2335423197492164, | |
| "grad_norm": 0.0595703125, | |
| "learning_rate": 3.208280522706602e-05, | |
| "loss": 0.0066, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 0.044677734375, | |
| "learning_rate": 2.9051371962893358e-05, | |
| "loss": 0.0063, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.311912225705329, | |
| "grad_norm": 0.05224609375, | |
| "learning_rate": 2.6145879860380773e-05, | |
| "loss": 0.0065, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.3510971786833856, | |
| "grad_norm": 0.0908203125, | |
| "learning_rate": 2.337148646888061e-05, | |
| "loss": 0.0065, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.3510971786833856, | |
| "eval_loss": 0.01185206975787878, | |
| "eval_runtime": 438.7429, | |
| "eval_samples_per_second": 2.587, | |
| "eval_steps_per_second": 1.295, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.3902821316614418, | |
| "grad_norm": 0.054931640625, | |
| "learning_rate": 2.073311662395764e-05, | |
| "loss": 0.0064, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 2.4294670846394983, | |
| "grad_norm": 0.0546875, | |
| "learning_rate": 1.8235453705295848e-05, | |
| "loss": 0.0063, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.468652037617555, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.5882931323214713e-05, | |
| "loss": 0.0065, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.507836990595611, | |
| "grad_norm": 0.072265625, | |
| "learning_rate": 1.3679725448551451e-05, | |
| "loss": 0.0062, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.5470219435736676, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 1.1629746999880697e-05, | |
| "loss": 0.0065, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "grad_norm": 0.0625, | |
| "learning_rate": 9.736634901228814e-06, | |
| "loss": 0.0064, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.6253918495297803, | |
| "grad_norm": 0.07177734375, | |
| "learning_rate": 8.00374962260706e-06, | |
| "loss": 0.0062, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 2.664576802507837, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 6.434167214829267e-06, | |
| "loss": 0.0062, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.7037617554858935, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 5.030673849203082e-06, | |
| "loss": 0.007, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 2.7429467084639496, | |
| "grad_norm": 0.11572265625, | |
| "learning_rate": 3.7957608717875015e-06, | |
| "loss": 0.0064, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.7429467084639496, | |
| "eval_loss": 0.011658398434519768, | |
| "eval_runtime": 438.6997, | |
| "eval_samples_per_second": 2.587, | |
| "eval_steps_per_second": 1.295, | |
| "step": 3500 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 3828, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 3.9128851808256e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |