| { | |
| "best_metric": 0.40765267610549927, | |
| "best_model_checkpoint": "m2m100_418M_finetuned_fr_to_sw/checkpoint-32000", | |
| "epoch": 5.977956286194657, | |
| "eval_steps": 1000, | |
| "global_step": 32000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 1.992527554642257e-05, | |
| "loss": 2.3549, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 1.9850551092845137e-05, | |
| "loss": 0.8843, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_bleu": 10.6171, | |
| "eval_gen_len": 60.1469, | |
| "eval_loss": 0.76387619972229, | |
| "eval_runtime": 3073.3727, | |
| "eval_samples_per_second": 3.483, | |
| "eval_steps_per_second": 0.436, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 1.97758266392677e-05, | |
| "loss": 0.7804, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 1.9701102185690268e-05, | |
| "loss": 0.7269, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_bleu": 14.3741, | |
| "eval_gen_len": 61.2398, | |
| "eval_loss": 0.6497731804847717, | |
| "eval_runtime": 3071.4883, | |
| "eval_samples_per_second": 3.486, | |
| "eval_steps_per_second": 0.436, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 1.9626377732112836e-05, | |
| "loss": 0.6685, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 1.9551653278535403e-05, | |
| "loss": 0.6504, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "eval_bleu": 19.1778, | |
| "eval_gen_len": 54.184, | |
| "eval_loss": 0.5995421409606934, | |
| "eval_runtime": 2745.8255, | |
| "eval_samples_per_second": 3.899, | |
| "eval_steps_per_second": 0.488, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.9476928824957967e-05, | |
| "loss": 0.6243, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 1.9402204371380535e-05, | |
| "loss": 0.6093, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_bleu": 20.3586, | |
| "eval_gen_len": 56.1132, | |
| "eval_loss": 0.5621405243873596, | |
| "eval_runtime": 2804.8241, | |
| "eval_samples_per_second": 3.817, | |
| "eval_steps_per_second": 0.477, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 1.9327479917803102e-05, | |
| "loss": 0.5779, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 1.925275546422567e-05, | |
| "loss": 0.58, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_bleu": 22.497, | |
| "eval_gen_len": 53.262, | |
| "eval_loss": 0.5317678451538086, | |
| "eval_runtime": 2510.0734, | |
| "eval_samples_per_second": 4.265, | |
| "eval_steps_per_second": 0.533, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 1.9178031010648237e-05, | |
| "loss": 0.5499, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 1.91033065570708e-05, | |
| "loss": 0.5067, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "eval_bleu": 24.1584, | |
| "eval_gen_len": 56.1712, | |
| "eval_loss": 0.5155890583992004, | |
| "eval_runtime": 2540.4899, | |
| "eval_samples_per_second": 4.214, | |
| "eval_steps_per_second": 0.527, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 1.902858210349337e-05, | |
| "loss": 0.5104, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 1.8953857649915936e-05, | |
| "loss": 0.4985, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "eval_bleu": 24.902, | |
| "eval_gen_len": 55.1034, | |
| "eval_loss": 0.5012524127960205, | |
| "eval_runtime": 2433.6643, | |
| "eval_samples_per_second": 4.399, | |
| "eval_steps_per_second": 0.55, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 1.8879133196338504e-05, | |
| "loss": 0.4949, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 1.880440874276107e-05, | |
| "loss": 0.4861, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "eval_bleu": 25.8945, | |
| "eval_gen_len": 55.7148, | |
| "eval_loss": 0.48973962664604187, | |
| "eval_runtime": 2476.4919, | |
| "eval_samples_per_second": 4.323, | |
| "eval_steps_per_second": 0.541, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 1.8729684289183636e-05, | |
| "loss": 0.4827, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 1.8654959835606203e-05, | |
| "loss": 0.4789, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "eval_bleu": 26.2593, | |
| "eval_gen_len": 54.9688, | |
| "eval_loss": 0.4776358902454376, | |
| "eval_runtime": 2500.3532, | |
| "eval_samples_per_second": 4.282, | |
| "eval_steps_per_second": 0.536, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 1.858023538202877e-05, | |
| "loss": 0.4757, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 1.8505510928451338e-05, | |
| "loss": 0.4748, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_bleu": 26.8308, | |
| "eval_gen_len": 53.234, | |
| "eval_loss": 0.4675232470035553, | |
| "eval_runtime": 2354.2254, | |
| "eval_samples_per_second": 4.548, | |
| "eval_steps_per_second": 0.569, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 1.8430786474873902e-05, | |
| "loss": 0.4721, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 1.835606202129647e-05, | |
| "loss": 0.4365, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_bleu": 27.8127, | |
| "eval_gen_len": 53.7894, | |
| "eval_loss": 0.46269142627716064, | |
| "eval_runtime": 2320.7276, | |
| "eval_samples_per_second": 4.613, | |
| "eval_steps_per_second": 0.577, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 1.8281337567719037e-05, | |
| "loss": 0.4124, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 1.8206613114141605e-05, | |
| "loss": 0.4065, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_bleu": 28.1334, | |
| "eval_gen_len": 52.4625, | |
| "eval_loss": 0.4552680253982544, | |
| "eval_runtime": 2260.3502, | |
| "eval_samples_per_second": 4.736, | |
| "eval_steps_per_second": 0.592, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 1.8131888660564172e-05, | |
| "loss": 0.4236, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 1.8057164206986736e-05, | |
| "loss": 0.4159, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "eval_bleu": 28.5473, | |
| "eval_gen_len": 53.1084, | |
| "eval_loss": 0.4502773582935333, | |
| "eval_runtime": 2305.3584, | |
| "eval_samples_per_second": 4.644, | |
| "eval_steps_per_second": 0.581, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 1.7982439753409304e-05, | |
| "loss": 0.4037, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 1.790771529983187e-05, | |
| "loss": 0.4078, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "eval_bleu": 28.522, | |
| "eval_gen_len": 53.9566, | |
| "eval_loss": 0.44360852241516113, | |
| "eval_runtime": 2357.175, | |
| "eval_samples_per_second": 4.542, | |
| "eval_steps_per_second": 0.568, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 1.783299084625444e-05, | |
| "loss": 0.4016, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 1.7758266392677006e-05, | |
| "loss": 0.4088, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_bleu": 29.6642, | |
| "eval_gen_len": 54.4689, | |
| "eval_loss": 0.439211368560791, | |
| "eval_runtime": 2379.609, | |
| "eval_samples_per_second": 4.499, | |
| "eval_steps_per_second": 0.563, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 1.768354193909957e-05, | |
| "loss": 0.4046, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 1.7608817485522138e-05, | |
| "loss": 0.4039, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_bleu": 29.8929, | |
| "eval_gen_len": 55.4612, | |
| "eval_loss": 0.4344358444213867, | |
| "eval_runtime": 2401.3922, | |
| "eval_samples_per_second": 4.458, | |
| "eval_steps_per_second": 0.558, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 1.7534093031944705e-05, | |
| "loss": 0.3635, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "learning_rate": 1.7459368578367273e-05, | |
| "loss": 0.3537, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "eval_bleu": 30.2302, | |
| "eval_gen_len": 54.3727, | |
| "eval_loss": 0.43423154950141907, | |
| "eval_runtime": 2351.6946, | |
| "eval_samples_per_second": 4.552, | |
| "eval_steps_per_second": 0.569, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 1.7384644124789837e-05, | |
| "loss": 0.3575, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "learning_rate": 1.7309919671212404e-05, | |
| "loss": 0.3569, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "eval_bleu": 30.1139, | |
| "eval_gen_len": 54.6381, | |
| "eval_loss": 0.4319211542606354, | |
| "eval_runtime": 2402.5651, | |
| "eval_samples_per_second": 4.456, | |
| "eval_steps_per_second": 0.557, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 1.7235195217634972e-05, | |
| "loss": 0.3564, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "learning_rate": 1.716047076405754e-05, | |
| "loss": 0.3564, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "eval_bleu": 30.8007, | |
| "eval_gen_len": 53.8819, | |
| "eval_loss": 0.42764702439308167, | |
| "eval_runtime": 2333.0447, | |
| "eval_samples_per_second": 4.589, | |
| "eval_steps_per_second": 0.574, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 1.7085746310480107e-05, | |
| "loss": 0.3576, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 1.701102185690267e-05, | |
| "loss": 0.3637, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "eval_bleu": 30.8698, | |
| "eval_gen_len": 53.7231, | |
| "eval_loss": 0.422607421875, | |
| "eval_runtime": 2331.5045, | |
| "eval_samples_per_second": 4.592, | |
| "eval_steps_per_second": 0.574, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 1.693629740332524e-05, | |
| "loss": 0.3601, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "learning_rate": 1.6861572949747806e-05, | |
| "loss": 0.3571, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "eval_bleu": 31.1343, | |
| "eval_gen_len": 53.5349, | |
| "eval_loss": 0.41751930117607117, | |
| "eval_runtime": 2304.4971, | |
| "eval_samples_per_second": 4.646, | |
| "eval_steps_per_second": 0.581, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "learning_rate": 1.6786848496170374e-05, | |
| "loss": 0.3441, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 1.671212404259294e-05, | |
| "loss": 0.3099, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "eval_bleu": 31.3026, | |
| "eval_gen_len": 53.4483, | |
| "eval_loss": 0.421342134475708, | |
| "eval_runtime": 2298.8454, | |
| "eval_samples_per_second": 4.657, | |
| "eval_steps_per_second": 0.582, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "learning_rate": 1.6637399589015505e-05, | |
| "loss": 0.3175, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "learning_rate": 1.6562675135438073e-05, | |
| "loss": 0.3104, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "eval_bleu": 31.1261, | |
| "eval_gen_len": 51.5196, | |
| "eval_loss": 0.4227532744407654, | |
| "eval_runtime": 2198.9363, | |
| "eval_samples_per_second": 4.869, | |
| "eval_steps_per_second": 0.609, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 1.648795068186064e-05, | |
| "loss": 0.3169, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 1.6413226228283208e-05, | |
| "loss": 0.3162, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "eval_bleu": 31.9091, | |
| "eval_gen_len": 53.0626, | |
| "eval_loss": 0.4195193946361542, | |
| "eval_runtime": 2270.3312, | |
| "eval_samples_per_second": 4.716, | |
| "eval_steps_per_second": 0.59, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "learning_rate": 1.6338501774705772e-05, | |
| "loss": 0.3128, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 1.626377732112834e-05, | |
| "loss": 0.3177, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "eval_bleu": 31.5561, | |
| "eval_gen_len": 52.3463, | |
| "eval_loss": 0.4158227741718292, | |
| "eval_runtime": 2237.9742, | |
| "eval_samples_per_second": 4.784, | |
| "eval_steps_per_second": 0.598, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 1.6189052867550907e-05, | |
| "loss": 0.3216, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 1.6114328413973474e-05, | |
| "loss": 0.3181, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "eval_bleu": 32.1029, | |
| "eval_gen_len": 53.8831, | |
| "eval_loss": 0.4130856692790985, | |
| "eval_runtime": 2288.7247, | |
| "eval_samples_per_second": 4.678, | |
| "eval_steps_per_second": 0.585, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "learning_rate": 1.6039603960396042e-05, | |
| "loss": 0.3176, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "learning_rate": 1.5964879506818606e-05, | |
| "loss": 0.2941, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "eval_bleu": 32.1061, | |
| "eval_gen_len": 52.7448, | |
| "eval_loss": 0.4150530993938446, | |
| "eval_runtime": 2247.1776, | |
| "eval_samples_per_second": 4.764, | |
| "eval_steps_per_second": 0.596, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "learning_rate": 1.5890155053241173e-05, | |
| "loss": 0.2752, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 5.23, | |
| "learning_rate": 1.581543059966374e-05, | |
| "loss": 0.274, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 5.23, | |
| "eval_bleu": 31.9128, | |
| "eval_gen_len": 52.9394, | |
| "eval_loss": 0.4146653711795807, | |
| "eval_runtime": 2267.9887, | |
| "eval_samples_per_second": 4.72, | |
| "eval_steps_per_second": 0.59, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "learning_rate": 1.574070614608631e-05, | |
| "loss": 0.2852, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "learning_rate": 1.5665981692508876e-05, | |
| "loss": 0.2713, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "eval_bleu": 32.452, | |
| "eval_gen_len": 52.881, | |
| "eval_loss": 0.41379648447036743, | |
| "eval_runtime": 2262.9812, | |
| "eval_samples_per_second": 4.731, | |
| "eval_steps_per_second": 0.592, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "learning_rate": 1.559125723893144e-05, | |
| "loss": 0.2791, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 1.5516532785354008e-05, | |
| "loss": 0.283, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "eval_bleu": 32.6103, | |
| "eval_gen_len": 53.2173, | |
| "eval_loss": 0.4103504717350006, | |
| "eval_runtime": 2279.2685, | |
| "eval_samples_per_second": 4.697, | |
| "eval_steps_per_second": 0.587, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "learning_rate": 1.5441808331776575e-05, | |
| "loss": 0.2835, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "learning_rate": 1.5367083878199142e-05, | |
| "loss": 0.2866, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 5.79, | |
| "eval_bleu": 32.5888, | |
| "eval_gen_len": 52.9638, | |
| "eval_loss": 0.41094180941581726, | |
| "eval_runtime": 2257.2287, | |
| "eval_samples_per_second": 4.743, | |
| "eval_steps_per_second": 0.593, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "learning_rate": 1.5292359424621707e-05, | |
| "loss": 0.2851, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "learning_rate": 1.5217634971044276e-05, | |
| "loss": 0.2865, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "eval_bleu": 32.6545, | |
| "eval_gen_len": 52.4693, | |
| "eval_loss": 0.40765267610549927, | |
| "eval_runtime": 2233.0208, | |
| "eval_samples_per_second": 4.794, | |
| "eval_steps_per_second": 0.6, | |
| "step": 32000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 133825, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 1000, | |
| "total_flos": 1.0234506177547469e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |