| { | |
| "best_metric": 1.1260286569595337, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-3_180-samples_config-2_full_auto/checkpoint-85", | |
| "epoch": 16.941176470588236, | |
| "eval_steps": 500, | |
| "global_step": 144, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 0.49722397327423096, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.6894, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 0.5352438688278198, | |
| "learning_rate": 5e-06, | |
| "loss": 1.7084, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.502581000328064, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6962, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 0.43059152364730835, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.6501, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.38872095942497253, | |
| "learning_rate": 2e-05, | |
| "loss": 1.674, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "eval_loss": 1.6583842039108276, | |
| "eval_runtime": 32.0223, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.5141124129295349, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.6671, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.35889914631843567, | |
| "learning_rate": 3e-05, | |
| "loss": 1.6062, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 0.42971694469451904, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.5823, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.4620816111564636, | |
| "learning_rate": 4e-05, | |
| "loss": 1.5596, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.5302298069000244, | |
| "eval_runtime": 32.0184, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 0.2591128349304199, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.5364, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.2332824170589447, | |
| "learning_rate": 5e-05, | |
| "loss": 1.4764, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.24145111441612244, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 1.4432, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.2435912787914276, | |
| "learning_rate": 6e-05, | |
| "loss": 1.4536, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 1.4046056270599365, | |
| "eval_runtime": 32.0313, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 0.23049457371234894, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.3867, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 0.27167701721191406, | |
| "learning_rate": 7e-05, | |
| "loss": 1.3341, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 0.2706312835216522, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.2763, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 0.2851117253303528, | |
| "learning_rate": 8e-05, | |
| "loss": 1.2837, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.24558836221694946, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.1955, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.239660382270813, | |
| "eval_runtime": 32.0211, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 0.21931374073028564, | |
| "learning_rate": 9e-05, | |
| "loss": 1.178, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 0.1845943033695221, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.1449, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 0.1530146449804306, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1571, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 0.15317296981811523, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 1.1221, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "eval_loss": 1.185463786125183, | |
| "eval_runtime": 32.0229, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 0.15968547761440277, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 1.0801, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 0.1755373626947403, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 1.1312, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 0.1728612631559372, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.0969, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 0.14794959127902985, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 1.1181, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.159245491027832, | |
| "eval_runtime": 32.0151, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 0.1490098237991333, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.0616, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 0.14090244472026825, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 1.0412, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 0.1525082290172577, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.0774, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 0.15122008323669434, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 1.1246, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.9411764705882355, | |
| "eval_loss": 1.1434063911437988, | |
| "eval_runtime": 32.0175, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 0.15296752750873566, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 1.0636, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 0.1479112207889557, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 1.0542, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 0.16554981470108032, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.0216, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 0.16862231492996216, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 1.0304, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.18343737721443176, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 1.0507, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.1335370540618896, | |
| "eval_runtime": 32.0137, | |
| "eval_samples_per_second": 1.125, | |
| "eval_steps_per_second": 1.125, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 0.17426958680152893, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 1.0151, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 0.19473035633563995, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.0459, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 0.19925150275230408, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 1.0283, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 0.20775443315505981, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.9763, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "eval_loss": 1.126503825187683, | |
| "eval_runtime": 32.0238, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 0.1641770899295807, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.9911, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 0.16879422962665558, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.0007, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 0.21175195276737213, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.9918, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 0.20613399147987366, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.9776, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.1260286569595337, | |
| "eval_runtime": 32.0216, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 0.2042946070432663, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.9517, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 0.21194055676460266, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.9831, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 0.21946534514427185, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.9295, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 0.2402062714099884, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.929, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "eval_loss": 1.1288561820983887, | |
| "eval_runtime": 32.013, | |
| "eval_samples_per_second": 1.125, | |
| "eval_steps_per_second": 1.125, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 11.058823529411764, | |
| "grad_norm": 0.23855681717395782, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 0.9745, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 11.294117647058824, | |
| "grad_norm": 0.2524814009666443, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.965, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 11.529411764705882, | |
| "grad_norm": 0.34287136793136597, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 0.8701, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 0.3116951882839203, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.9165, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.3116912543773651, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 0.8679, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.1547869443893433, | |
| "eval_runtime": 32.0171, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.235294117647058, | |
| "grad_norm": 0.3423101305961609, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.8385, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 12.470588235294118, | |
| "grad_norm": 0.3736465871334076, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 0.847, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 12.705882352941176, | |
| "grad_norm": 0.39747124910354614, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 0.8773, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "grad_norm": 0.40393248200416565, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.8616, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "eval_loss": 1.168959617614746, | |
| "eval_runtime": 32.0142, | |
| "eval_samples_per_second": 1.125, | |
| "eval_steps_per_second": 1.125, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 13.176470588235293, | |
| "grad_norm": 0.3690202832221985, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.8302, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 13.411764705882353, | |
| "grad_norm": 0.4534393548965454, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 0.7597, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 13.647058823529411, | |
| "grad_norm": 0.47646868228912354, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 0.7924, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 13.882352941176471, | |
| "grad_norm": 0.5564125776290894, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 0.815, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.206106424331665, | |
| "eval_runtime": 32.0191, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 14.117647058823529, | |
| "grad_norm": 0.5058925151824951, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.7954, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 14.352941176470589, | |
| "grad_norm": 0.5072349905967712, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 0.6638, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 14.588235294117647, | |
| "grad_norm": 0.5933027863502502, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.7629, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 14.823529411764707, | |
| "grad_norm": 0.6190547347068787, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 0.6912, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 14.941176470588236, | |
| "eval_loss": 1.2299973964691162, | |
| "eval_runtime": 32.0138, | |
| "eval_samples_per_second": 1.125, | |
| "eval_steps_per_second": 1.125, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 15.058823529411764, | |
| "grad_norm": 0.6815229654312134, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.7476, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 15.294117647058824, | |
| "grad_norm": 0.987908661365509, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.66, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 15.529411764705882, | |
| "grad_norm": 0.791010856628418, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 0.6825, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 15.764705882352942, | |
| "grad_norm": 0.9056291580200195, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 0.6842, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.7855310440063477, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.6046, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.295262098312378, | |
| "eval_runtime": 32.0163, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 16.235294117647058, | |
| "grad_norm": 3.381535530090332, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 0.5623, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 16.470588235294116, | |
| "grad_norm": 0.8216532468795776, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.6597, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 16.705882352941178, | |
| "grad_norm": 0.8799306154251099, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 0.5784, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 16.941176470588236, | |
| "grad_norm": 1.140483021736145, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.5919, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 16.941176470588236, | |
| "eval_loss": 1.3360047340393066, | |
| "eval_runtime": 32.0291, | |
| "eval_samples_per_second": 1.124, | |
| "eval_steps_per_second": 1.124, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 16.941176470588236, | |
| "step": 144, | |
| "total_flos": 1.826550570853335e+17, | |
| "train_loss": 1.0546488515618775, | |
| "train_runtime": 6693.3246, | |
| "train_samples_per_second": 1.016, | |
| "train_steps_per_second": 0.06 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.826550570853335e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |