| { | |
| "best_metric": 1.3743919134140015, | |
| "best_model_checkpoint": "./llama3/21-04-24-Weni-WeniGPT-Agents-Llama3-1.0.8-SFT_Experiment with SFT and Llama3 and updates in requirements-2_max_steps-669_batch_2_2024-04-21_ppid_2917/checkpoint-360", | |
| "epoch": 2.4161073825503356, | |
| "eval_steps": 30, | |
| "global_step": 540, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0447427293064877, | |
| "grad_norm": 0.3199349641799927, | |
| "learning_rate": 0.0001, | |
| "loss": 1.9361, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0894854586129754, | |
| "grad_norm": 0.9833651781082153, | |
| "learning_rate": 0.0002, | |
| "loss": 1.8397, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1342281879194631, | |
| "grad_norm": 1.0959680080413818, | |
| "learning_rate": 0.00019990511519199923, | |
| "loss": 1.7877, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1342281879194631, | |
| "eval_loss": 1.7081660032272339, | |
| "eval_runtime": 10.9642, | |
| "eval_samples_per_second": 4.195, | |
| "eval_steps_per_second": 4.195, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1789709172259508, | |
| "grad_norm": 1.2613288164138794, | |
| "learning_rate": 0.00019957734949126304, | |
| "loss": 1.7523, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22371364653243847, | |
| "grad_norm": 0.7975640892982483, | |
| "learning_rate": 0.0001990162991697884, | |
| "loss": 1.516, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2684563758389262, | |
| "grad_norm": 0.71113121509552, | |
| "learning_rate": 0.0001982232786270059, | |
| "loss": 1.3675, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2684563758389262, | |
| "eval_loss": 1.4538882970809937, | |
| "eval_runtime": 10.9625, | |
| "eval_samples_per_second": 4.196, | |
| "eval_steps_per_second": 4.196, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3131991051454139, | |
| "grad_norm": 0.8126013875007629, | |
| "learning_rate": 0.00019720014571008158, | |
| "loss": 1.4173, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3579418344519016, | |
| "grad_norm": 0.6399104595184326, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 1.5249, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.40268456375838924, | |
| "grad_norm": 0.7685297131538391, | |
| "learning_rate": 0.00019447366400338116, | |
| "loss": 1.3782, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.40268456375838924, | |
| "eval_loss": 1.4235661029815674, | |
| "eval_runtime": 10.9616, | |
| "eval_samples_per_second": 4.196, | |
| "eval_steps_per_second": 4.196, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.44742729306487694, | |
| "grad_norm": 0.7407508492469788, | |
| "learning_rate": 0.00019277670267274258, | |
| "loss": 1.3854, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.49217002237136465, | |
| "grad_norm": 0.5381172895431519, | |
| "learning_rate": 0.0001908623889220311, | |
| "loss": 1.3216, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5369127516778524, | |
| "grad_norm": 0.8507822155952454, | |
| "learning_rate": 0.00018873520750565718, | |
| "loss": 1.3884, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5369127516778524, | |
| "eval_loss": 1.3937652111053467, | |
| "eval_runtime": 11.3392, | |
| "eval_samples_per_second": 4.057, | |
| "eval_steps_per_second": 4.057, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5816554809843401, | |
| "grad_norm": 0.7832561135292053, | |
| "learning_rate": 0.00018640014187329578, | |
| "loss": 1.3304, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6263982102908278, | |
| "grad_norm": 0.7797898054122925, | |
| "learning_rate": 0.00018386266249492057, | |
| "loss": 1.4979, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6711409395973155, | |
| "grad_norm": 0.8133324384689331, | |
| "learning_rate": 0.00018112871404487202, | |
| "loss": 1.3448, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6711409395973155, | |
| "eval_loss": 1.3898664712905884, | |
| "eval_runtime": 10.9657, | |
| "eval_samples_per_second": 4.195, | |
| "eval_steps_per_second": 4.195, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7158836689038032, | |
| "grad_norm": 0.5568630695343018, | |
| "learning_rate": 0.00017820470147498455, | |
| "loss": 1.3767, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7606263982102909, | |
| "grad_norm": 0.6193764209747314, | |
| "learning_rate": 0.00017509747500939928, | |
| "loss": 1.3777, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.8053691275167785, | |
| "grad_norm": 0.8501898646354675, | |
| "learning_rate": 0.00017181431409621644, | |
| "loss": 1.3357, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8053691275167785, | |
| "eval_loss": 1.3869752883911133, | |
| "eval_runtime": 10.9836, | |
| "eval_samples_per_second": 4.188, | |
| "eval_steps_per_second": 4.188, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8501118568232662, | |
| "grad_norm": 0.6570385098457336, | |
| "learning_rate": 0.00016836291035358375, | |
| "loss": 1.3534, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8948545861297539, | |
| "grad_norm": 0.8461380004882812, | |
| "learning_rate": 0.0001647513495501749, | |
| "loss": 1.3976, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9395973154362416, | |
| "grad_norm": 0.9425082802772522, | |
| "learning_rate": 0.000160988092662272, | |
| "loss": 1.2788, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9395973154362416, | |
| "eval_loss": 1.3739150762557983, | |
| "eval_runtime": 10.967, | |
| "eval_samples_per_second": 4.194, | |
| "eval_steps_per_second": 4.194, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9843400447427293, | |
| "grad_norm": 1.4384957551956177, | |
| "learning_rate": 0.0001570819560518322, | |
| "loss": 1.3569, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.029082774049217, | |
| "grad_norm": 0.665977418422699, | |
| "learning_rate": 0.00015304209081197425, | |
| "loss": 1.3895, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0738255033557047, | |
| "grad_norm": 0.736944317817688, | |
| "learning_rate": 0.0001488779613282751, | |
| "loss": 1.2396, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0738255033557047, | |
| "eval_loss": 1.3777718544006348, | |
| "eval_runtime": 10.9845, | |
| "eval_samples_per_second": 4.188, | |
| "eval_steps_per_second": 4.188, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.1185682326621924, | |
| "grad_norm": 0.5746013522148132, | |
| "learning_rate": 0.00014459932310610093, | |
| "loss": 1.3425, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.1633109619686801, | |
| "grad_norm": 0.8226341009140015, | |
| "learning_rate": 0.00014021619991591794, | |
| "loss": 1.2856, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.2080536912751678, | |
| "grad_norm": 1.051847219467163, | |
| "learning_rate": 0.00013573886031012584, | |
| "loss": 1.2949, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.2080536912751678, | |
| "eval_loss": 1.3808656930923462, | |
| "eval_runtime": 10.966, | |
| "eval_samples_per_second": 4.195, | |
| "eval_steps_per_second": 4.195, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.2527964205816555, | |
| "grad_norm": 0.6205160021781921, | |
| "learning_rate": 0.00013117779356642872, | |
| "loss": 1.2785, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.2975391498881432, | |
| "grad_norm": 1.1361136436462402, | |
| "learning_rate": 0.00012654368511410245, | |
| "loss": 1.2825, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.342281879194631, | |
| "grad_norm": 0.8586457967758179, | |
| "learning_rate": 0.00012184739150072821, | |
| "loss": 1.337, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.342281879194631, | |
| "eval_loss": 1.3791558742523193, | |
| "eval_runtime": 12.241, | |
| "eval_samples_per_second": 3.758, | |
| "eval_steps_per_second": 3.758, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.3870246085011186, | |
| "grad_norm": 1.0672543048858643, | |
| "learning_rate": 0.00011709991495803915, | |
| "loss": 1.1912, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.4317673378076063, | |
| "grad_norm": 0.8796883821487427, | |
| "learning_rate": 0.0001123123776264656, | |
| "loss": 1.1901, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.476510067114094, | |
| "grad_norm": 0.9798341989517212, | |
| "learning_rate": 0.00010749599549876472, | |
| "loss": 1.3266, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.476510067114094, | |
| "eval_loss": 1.3774683475494385, | |
| "eval_runtime": 10.9641, | |
| "eval_samples_per_second": 4.196, | |
| "eval_steps_per_second": 4.196, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.5212527964205815, | |
| "grad_norm": 1.070554494857788, | |
| "learning_rate": 0.00010266205214377748, | |
| "loss": 1.255, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.5659955257270695, | |
| "grad_norm": 0.8746442794799805, | |
| "learning_rate": 9.782187227187231e-05, | |
| "loss": 1.1981, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.610738255033557, | |
| "grad_norm": 1.1643601655960083, | |
| "learning_rate": 9.298679520400412e-05, | |
| "loss": 1.2735, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.610738255033557, | |
| "eval_loss": 1.3743919134140015, | |
| "eval_runtime": 10.9721, | |
| "eval_samples_per_second": 4.192, | |
| "eval_steps_per_second": 4.192, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.6554809843400449, | |
| "grad_norm": 1.2565709352493286, | |
| "learning_rate": 8.816814830654468e-05, | |
| "loss": 1.1877, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.7002237136465324, | |
| "grad_norm": 1.3815152645111084, | |
| "learning_rate": 8.33772204541195e-05, | |
| "loss": 1.2719, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.7449664429530203, | |
| "grad_norm": 1.1193474531173706, | |
| "learning_rate": 7.862523558262116e-05, | |
| "loss": 1.2809, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.7449664429530203, | |
| "eval_loss": 1.37523353099823, | |
| "eval_runtime": 12.3393, | |
| "eval_samples_per_second": 3.728, | |
| "eval_steps_per_second": 3.728, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.7897091722595078, | |
| "grad_norm": 0.9928935766220093, | |
| "learning_rate": 7.392332639435752e-05, | |
| "loss": 1.3729, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.8344519015659957, | |
| "grad_norm": 1.6183288097381592, | |
| "learning_rate": 6.928250827693771e-05, | |
| "loss": 1.1143, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.8791946308724832, | |
| "grad_norm": 1.1342641115188599, | |
| "learning_rate": 6.471365349699636e-05, | |
| "loss": 1.2383, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.8791946308724832, | |
| "eval_loss": 1.377549648284912, | |
| "eval_runtime": 10.9609, | |
| "eval_samples_per_second": 4.197, | |
| "eval_steps_per_second": 4.197, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.9239373601789709, | |
| "grad_norm": 0.9746946096420288, | |
| "learning_rate": 6.022746572921447e-05, | |
| "loss": 1.1807, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.9686800894854586, | |
| "grad_norm": 1.2227619886398315, | |
| "learning_rate": 5.583445498030848e-05, | |
| "loss": 1.1108, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.0134228187919465, | |
| "grad_norm": 1.0090824365615845, | |
| "learning_rate": 5.1544912966734994e-05, | |
| "loss": 1.2116, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.0134228187919465, | |
| "eval_loss": 1.385948657989502, | |
| "eval_runtime": 10.9657, | |
| "eval_samples_per_second": 4.195, | |
| "eval_steps_per_second": 4.195, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.058165548098434, | |
| "grad_norm": 1.5286532640457153, | |
| "learning_rate": 4.7368889003794026e-05, | |
| "loss": 1.1827, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.1029082774049215, | |
| "grad_norm": 1.072304368019104, | |
| "learning_rate": 4.3316166462617355e-05, | |
| "loss": 1.2313, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.1476510067114094, | |
| "grad_norm": 1.698246717453003, | |
| "learning_rate": 3.939623985019679e-05, | |
| "loss": 1.0153, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.1476510067114094, | |
| "eval_loss": 1.3925849199295044, | |
| "eval_runtime": 10.9662, | |
| "eval_samples_per_second": 4.195, | |
| "eval_steps_per_second": 4.195, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.192393736017897, | |
| "grad_norm": 1.3146202564239502, | |
| "learning_rate": 3.561829256614856e-05, | |
| "loss": 1.2576, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.237136465324385, | |
| "grad_norm": 1.545972466468811, | |
| "learning_rate": 3.199117538832358e-05, | |
| "loss": 1.2075, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.2818791946308723, | |
| "grad_norm": 1.2800744771957397, | |
| "learning_rate": 2.852338573766675e-05, | |
| "loss": 1.2039, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.2818791946308723, | |
| "eval_loss": 1.3884193897247314, | |
| "eval_runtime": 10.9618, | |
| "eval_samples_per_second": 4.196, | |
| "eval_steps_per_second": 4.196, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.3266219239373602, | |
| "grad_norm": 1.2421742677688599, | |
| "learning_rate": 2.5223047770902274e-05, | |
| "loss": 1.298, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.3713646532438477, | |
| "grad_norm": 0.9423472285270691, | |
| "learning_rate": 2.2097893347683198e-05, | |
| "loss": 1.1545, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.4161073825503356, | |
| "grad_norm": 1.5044140815734863, | |
| "learning_rate": 1.915524391679375e-05, | |
| "loss": 1.2451, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.4161073825503356, | |
| "eval_loss": 1.388563871383667, | |
| "eval_runtime": 10.9674, | |
| "eval_samples_per_second": 4.194, | |
| "eval_steps_per_second": 4.194, | |
| "step": 540 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 669, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 90, | |
| "total_flos": 1.1001422075019264e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |