| { | |
| "best_metric": 0.09502185136079788, | |
| "best_model_checkpoint": "fine-tuning-Phi2-with-webglm-qa-with-lora_7/checkpoint-1000", | |
| "epoch": 15.723270440251572, | |
| "eval_steps": 20, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 7.3505, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_loss": 6.286308765411377, | |
| "eval_runtime": 25.4764, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 4.0914, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_loss": 0.9255164861679077, | |
| "eval_runtime": 25.461, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6517, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_loss": 0.576193630695343, | |
| "eval_runtime": 25.4995, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.431, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 4.893617021276596e-05, | |
| "loss": 0.4621, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_loss": 0.40622055530548096, | |
| "eval_runtime": 25.4724, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 4.787234042553192e-05, | |
| "loss": 0.3128, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "eval_loss": 0.30563685297966003, | |
| "eval_runtime": 25.4374, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 4.680851063829788e-05, | |
| "loss": 0.2536, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "eval_loss": 0.2603812515735626, | |
| "eval_runtime": 25.4763, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 4.574468085106383e-05, | |
| "loss": 0.2227, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_loss": 0.22473430633544922, | |
| "eval_runtime": 25.4963, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.431, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 4.468085106382979e-05, | |
| "loss": 0.1901, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "eval_loss": 0.2041008621454239, | |
| "eval_runtime": 25.459, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 4.3617021276595746e-05, | |
| "loss": 0.176, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "eval_loss": 0.1812312752008438, | |
| "eval_runtime": 25.4868, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.432, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 4.2553191489361704e-05, | |
| "loss": 0.1453, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "eval_loss": 0.16830970346927643, | |
| "eval_runtime": 25.481, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.432, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 4.148936170212766e-05, | |
| "loss": 0.1557, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "eval_loss": 0.15921901166439056, | |
| "eval_runtime": 25.5028, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.431, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 4.0425531914893614e-05, | |
| "loss": 0.1441, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "eval_loss": 0.14882861077785492, | |
| "eval_runtime": 25.4427, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "learning_rate": 3.936170212765958e-05, | |
| "loss": 0.1282, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "eval_loss": 0.1429978758096695, | |
| "eval_runtime": 25.4905, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.432, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 3.829787234042553e-05, | |
| "loss": 0.1215, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "eval_loss": 0.13476386666297913, | |
| "eval_runtime": 25.4804, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.432, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 3.723404255319149e-05, | |
| "loss": 0.1217, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "eval_loss": 0.13230670988559723, | |
| "eval_runtime": 25.4706, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "learning_rate": 3.617021276595745e-05, | |
| "loss": 0.117, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "eval_loss": 0.12707628309726715, | |
| "eval_runtime": 25.4283, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.433, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "learning_rate": 3.5106382978723407e-05, | |
| "loss": 0.109, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.35, | |
| "eval_loss": 0.12554581463336945, | |
| "eval_runtime": 25.4583, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "learning_rate": 3.4042553191489365e-05, | |
| "loss": 0.1094, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "eval_loss": 0.12099047005176544, | |
| "eval_runtime": 25.4606, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.97, | |
| "learning_rate": 3.2978723404255317e-05, | |
| "loss": 0.1057, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 5.97, | |
| "eval_loss": 0.11747618019580841, | |
| "eval_runtime": 25.4382, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "learning_rate": 3.191489361702128e-05, | |
| "loss": 0.0937, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.29, | |
| "eval_loss": 0.11580007523298264, | |
| "eval_runtime": 25.4149, | |
| "eval_samples_per_second": 0.866, | |
| "eval_steps_per_second": 0.433, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 3.085106382978723e-05, | |
| "loss": 0.0942, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "eval_loss": 0.11587951332330704, | |
| "eval_runtime": 25.4544, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.92, | |
| "learning_rate": 2.9787234042553192e-05, | |
| "loss": 0.1007, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 6.92, | |
| "eval_loss": 0.11250220984220505, | |
| "eval_runtime": 25.4316, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.433, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "learning_rate": 2.8723404255319154e-05, | |
| "loss": 0.0876, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "eval_loss": 0.11185076832771301, | |
| "eval_runtime": 25.4765, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "learning_rate": 2.765957446808511e-05, | |
| "loss": 0.0894, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "eval_loss": 0.1098945215344429, | |
| "eval_runtime": 25.4185, | |
| "eval_samples_per_second": 0.866, | |
| "eval_steps_per_second": 0.433, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "learning_rate": 2.6595744680851064e-05, | |
| "loss": 0.0827, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.86, | |
| "eval_loss": 0.10724210739135742, | |
| "eval_runtime": 25.4969, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.431, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.18, | |
| "learning_rate": 2.5531914893617022e-05, | |
| "loss": 0.0894, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 8.18, | |
| "eval_loss": 0.10687847435474396, | |
| "eval_runtime": 25.4493, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "learning_rate": 2.446808510638298e-05, | |
| "loss": 0.0805, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "eval_loss": 0.10752053558826447, | |
| "eval_runtime": 25.4419, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 8.81, | |
| "learning_rate": 2.340425531914894e-05, | |
| "loss": 0.0782, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 8.81, | |
| "eval_loss": 0.10425343364477158, | |
| "eval_runtime": 25.4865, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.432, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 9.12, | |
| "learning_rate": 2.2340425531914894e-05, | |
| "loss": 0.0881, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 9.12, | |
| "eval_loss": 0.1033911183476448, | |
| "eval_runtime": 25.4549, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "learning_rate": 2.1276595744680852e-05, | |
| "loss": 0.0839, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "eval_loss": 0.10145573318004608, | |
| "eval_runtime": 25.4481, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.75, | |
| "learning_rate": 2.0212765957446807e-05, | |
| "loss": 0.0694, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 9.75, | |
| "eval_loss": 0.10002648830413818, | |
| "eval_runtime": 25.4089, | |
| "eval_samples_per_second": 0.866, | |
| "eval_steps_per_second": 0.433, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "learning_rate": 1.9148936170212766e-05, | |
| "loss": 0.068, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "eval_loss": 0.10073428601026535, | |
| "eval_runtime": 25.4963, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.431, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 10.38, | |
| "learning_rate": 1.8085106382978724e-05, | |
| "loss": 0.072, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 10.38, | |
| "eval_loss": 0.09940142929553986, | |
| "eval_runtime": 25.4629, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 10.69, | |
| "learning_rate": 1.7021276595744682e-05, | |
| "loss": 0.0709, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 10.69, | |
| "eval_loss": 0.09847569465637207, | |
| "eval_runtime": 25.5056, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.431, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "learning_rate": 1.595744680851064e-05, | |
| "loss": 0.0712, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "eval_loss": 0.09858354926109314, | |
| "eval_runtime": 25.4627, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 11.32, | |
| "learning_rate": 1.4893617021276596e-05, | |
| "loss": 0.0673, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 11.32, | |
| "eval_loss": 0.09991483390331268, | |
| "eval_runtime": 25.4891, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.432, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 11.64, | |
| "learning_rate": 1.3829787234042554e-05, | |
| "loss": 0.0669, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 11.64, | |
| "eval_loss": 0.09741941094398499, | |
| "eval_runtime": 25.4606, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 11.95, | |
| "learning_rate": 1.2765957446808511e-05, | |
| "loss": 0.0706, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 11.95, | |
| "eval_loss": 0.0980500727891922, | |
| "eval_runtime": 25.4509, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "learning_rate": 1.170212765957447e-05, | |
| "loss": 0.0641, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "eval_loss": 0.09693228453397751, | |
| "eval_runtime": 25.436, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 12.58, | |
| "learning_rate": 1.0638297872340426e-05, | |
| "loss": 0.0652, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.58, | |
| "eval_loss": 0.0963829830288887, | |
| "eval_runtime": 25.4744, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 12.89, | |
| "learning_rate": 9.574468085106383e-06, | |
| "loss": 0.0668, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 12.89, | |
| "eval_loss": 0.09619712829589844, | |
| "eval_runtime": 25.446, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 13.21, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.0617, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 13.21, | |
| "eval_loss": 0.09718295931816101, | |
| "eval_runtime": 25.4457, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 13.52, | |
| "learning_rate": 7.446808510638298e-06, | |
| "loss": 0.0628, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 13.52, | |
| "eval_loss": 0.09600641578435898, | |
| "eval_runtime": 25.4516, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 13.84, | |
| "learning_rate": 6.3829787234042555e-06, | |
| "loss": 0.0637, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 13.84, | |
| "eval_loss": 0.09490146487951279, | |
| "eval_runtime": 25.4889, | |
| "eval_samples_per_second": 0.863, | |
| "eval_steps_per_second": 0.432, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 14.15, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 0.0633, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 14.15, | |
| "eval_loss": 0.0950675681233406, | |
| "eval_runtime": 25.4561, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 14.47, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.0577, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 14.47, | |
| "eval_loss": 0.09526454657316208, | |
| "eval_runtime": 25.5288, | |
| "eval_samples_per_second": 0.862, | |
| "eval_steps_per_second": 0.431, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 14.78, | |
| "learning_rate": 3.1914893617021277e-06, | |
| "loss": 0.0646, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 14.78, | |
| "eval_loss": 0.09467268735170364, | |
| "eval_runtime": 25.4341, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 15.09, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 0.06, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 15.09, | |
| "eval_loss": 0.09462392330169678, | |
| "eval_runtime": 25.446, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 15.41, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "loss": 0.0584, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 15.41, | |
| "eval_loss": 0.09491446614265442, | |
| "eval_runtime": 25.4603, | |
| "eval_samples_per_second": 0.864, | |
| "eval_steps_per_second": 0.432, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 15.72, | |
| "learning_rate": 0.0, | |
| "loss": 0.0638, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 15.72, | |
| "eval_loss": 0.09502185136079788, | |
| "eval_runtime": 25.4343, | |
| "eval_samples_per_second": 0.865, | |
| "eval_steps_per_second": 0.432, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 1000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 16, | |
| "save_steps": 200, | |
| "total_flos": 1.63933992517632e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |