| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 6, | |
| "global_step": 78, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.038461538461538464, | |
| "grad_norm": 8.71969985961914, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7925, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.038461538461538464, | |
| "eval_loss": 2.0411531925201416, | |
| "eval_runtime": 1.6566, | |
| "eval_samples_per_second": 13.28, | |
| "eval_steps_per_second": 1.207, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 4.587398052215576, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1894, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11538461538461539, | |
| "grad_norm": 4.494429111480713, | |
| "learning_rate": 3e-05, | |
| "loss": 1.9438, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 7.78571891784668, | |
| "learning_rate": 4e-05, | |
| "loss": 1.9463, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 6.792930603027344, | |
| "learning_rate": 5e-05, | |
| "loss": 1.8652, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 10.765946388244629, | |
| "learning_rate": 6e-05, | |
| "loss": 1.6872, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "eval_loss": 1.6089171171188354, | |
| "eval_runtime": 1.667, | |
| "eval_samples_per_second": 13.198, | |
| "eval_steps_per_second": 1.2, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2692307692307692, | |
| "grad_norm": 3.0110535621643066, | |
| "learning_rate": 7e-05, | |
| "loss": 1.6877, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.871809959411621, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4855, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.34615384615384615, | |
| "grad_norm": 2.223036050796509, | |
| "learning_rate": 9e-05, | |
| "loss": 1.2108, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 2.2891507148742676, | |
| "learning_rate": 0.0001, | |
| "loss": 1.0259, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4230769230769231, | |
| "grad_norm": 2.058546543121338, | |
| "learning_rate": 9.997207818651274e-05, | |
| "loss": 0.8116, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 2.8107495307922363, | |
| "learning_rate": 9.988834393115767e-05, | |
| "loss": 0.6967, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "eval_loss": 0.6327503323554993, | |
| "eval_runtime": 1.6315, | |
| "eval_samples_per_second": 13.485, | |
| "eval_steps_per_second": 1.226, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.0721328258514404, | |
| "learning_rate": 9.974889075442521e-05, | |
| "loss": 0.5983, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 1.9303380250930786, | |
| "learning_rate": 9.9553874407739e-05, | |
| "loss": 0.4963, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 1.40817391872406, | |
| "learning_rate": 9.930351269950143e-05, | |
| "loss": 0.3881, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.4169831275939941, | |
| "learning_rate": 9.899808525182935e-05, | |
| "loss": 0.3139, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.6538461538461539, | |
| "grad_norm": 0.9847315549850464, | |
| "learning_rate": 9.863793318825186e-05, | |
| "loss": 0.2776, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 1.2575572729110718, | |
| "learning_rate": 9.822345875271883e-05, | |
| "loss": 0.3327, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "eval_loss": 0.2710803747177124, | |
| "eval_runtime": 1.6739, | |
| "eval_samples_per_second": 13.143, | |
| "eval_steps_per_second": 1.195, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7307692307692307, | |
| "grad_norm": 0.7373743057250977, | |
| "learning_rate": 9.775512486034563e-05, | |
| "loss": 0.1967, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.8841055035591125, | |
| "learning_rate": 9.723345458039594e-05, | |
| "loss": 0.244, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8076923076923077, | |
| "grad_norm": 0.8621382713317871, | |
| "learning_rate": 9.665903055208014e-05, | |
| "loss": 0.1927, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.9871663451194763, | |
| "learning_rate": 9.603249433382144e-05, | |
| "loss": 0.2435, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.8846153846153846, | |
| "grad_norm": 0.6780880689620972, | |
| "learning_rate": 9.535454568671704e-05, | |
| "loss": 0.1498, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.7006078958511353, | |
| "learning_rate": 9.462594179299406e-05, | |
| "loss": 0.1784, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "eval_loss": 0.17325076460838318, | |
| "eval_runtime": 1.6307, | |
| "eval_samples_per_second": 13.491, | |
| "eval_steps_per_second": 1.226, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 0.69418865442276, | |
| "learning_rate": 9.384749641033359e-05, | |
| "loss": 0.2146, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5629783272743225, | |
| "learning_rate": 9.302007896300698e-05, | |
| "loss": 0.1474, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0384615384615385, | |
| "grad_norm": 0.5029877424240112, | |
| "learning_rate": 9.214461357083985e-05, | |
| "loss": 0.1358, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.5899357199668884, | |
| "learning_rate": 9.122207801708802e-05, | |
| "loss": 0.134, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.1153846153846154, | |
| "grad_norm": 0.7418510913848877, | |
| "learning_rate": 9.025350265637815e-05, | |
| "loss": 0.1158, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 0.7511109113693237, | |
| "learning_rate": 8.923996926393305e-05, | |
| "loss": 0.1136, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "eval_loss": 0.1189715564250946, | |
| "eval_runtime": 1.672, | |
| "eval_samples_per_second": 13.158, | |
| "eval_steps_per_second": 1.196, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1923076923076923, | |
| "grad_norm": 0.7893090844154358, | |
| "learning_rate": 8.818260982736661e-05, | |
| "loss": 0.1249, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.5709223747253418, | |
| "learning_rate": 8.708260528239788e-05, | |
| "loss": 0.0824, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.2692307692307692, | |
| "grad_norm": 0.5314195156097412, | |
| "learning_rate": 8.594118419389647e-05, | |
| "loss": 0.0906, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.3076923076923077, | |
| "grad_norm": 0.5469984412193298, | |
| "learning_rate": 8.475962138373213e-05, | |
| "loss": 0.0918, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3461538461538463, | |
| "grad_norm": 0.5641104578971863, | |
| "learning_rate": 8.353923650696118e-05, | |
| "loss": 0.0866, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.5818964838981628, | |
| "learning_rate": 8.228139257794012e-05, | |
| "loss": 0.0891, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "eval_loss": 0.08501879870891571, | |
| "eval_runtime": 1.6676, | |
| "eval_samples_per_second": 13.193, | |
| "eval_steps_per_second": 1.199, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4230769230769231, | |
| "grad_norm": 0.7663794159889221, | |
| "learning_rate": 8.098749444801224e-05, | |
| "loss": 0.0934, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.4615384615384617, | |
| "grad_norm": 0.48853516578674316, | |
| "learning_rate": 7.965898723646776e-05, | |
| "loss": 0.0597, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.608406662940979, | |
| "learning_rate": 7.829735471652978e-05, | |
| "loss": 0.0579, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.49656111001968384, | |
| "learning_rate": 7.690411765816864e-05, | |
| "loss": 0.0607, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5769230769230769, | |
| "grad_norm": 0.6709762215614319, | |
| "learning_rate": 7.548083212959588e-05, | |
| "loss": 0.0761, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "grad_norm": 0.6458861827850342, | |
| "learning_rate": 7.402908775933419e-05, | |
| "loss": 0.0746, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6153846153846154, | |
| "eval_loss": 0.06257763504981995, | |
| "eval_runtime": 1.6397, | |
| "eval_samples_per_second": 13.417, | |
| "eval_steps_per_second": 1.22, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6538461538461537, | |
| "grad_norm": 0.527205765247345, | |
| "learning_rate": 7.255050596080509e-05, | |
| "loss": 0.0597, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.4971027672290802, | |
| "learning_rate": 7.104673812141675e-05, | |
| "loss": 0.0574, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7307692307692308, | |
| "grad_norm": 0.521812915802002, | |
| "learning_rate": 6.951946375817474e-05, | |
| "loss": 0.0611, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7692307692307692, | |
| "grad_norm": 0.49732235074043274, | |
| "learning_rate": 6.797038864187564e-05, | |
| "loss": 0.0411, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8076923076923077, | |
| "grad_norm": 0.4331294000148773, | |
| "learning_rate": 6.640124289197845e-05, | |
| "loss": 0.0377, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.5265694260597229, | |
| "learning_rate": 6.481377904428171e-05, | |
| "loss": 0.0522, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "eval_loss": 0.046460047364234924, | |
| "eval_runtime": 1.6754, | |
| "eval_samples_per_second": 13.131, | |
| "eval_steps_per_second": 1.194, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.8846153846153846, | |
| "grad_norm": 0.5977218747138977, | |
| "learning_rate": 6.320977009356431e-05, | |
| "loss": 0.0543, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 0.4617597758769989, | |
| "learning_rate": 6.159100751337642e-05, | |
| "loss": 0.0388, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9615384615384617, | |
| "grad_norm": 0.4888545572757721, | |
| "learning_rate": 5.99592992551918e-05, | |
| "loss": 0.0377, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.43416959047317505, | |
| "learning_rate": 5.831646772915651e-05, | |
| "loss": 0.0326, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.0384615384615383, | |
| "grad_norm": 0.4446072578430176, | |
| "learning_rate": 5.666434776868895e-05, | |
| "loss": 0.03, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "grad_norm": 0.569553792476654, | |
| "learning_rate": 5.5004784581204927e-05, | |
| "loss": 0.033, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.076923076923077, | |
| "eval_loss": 0.028210198506712914, | |
| "eval_runtime": 1.7415, | |
| "eval_samples_per_second": 12.633, | |
| "eval_steps_per_second": 1.148, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1153846153846154, | |
| "grad_norm": 0.4060160219669342, | |
| "learning_rate": 5.3339631687256084e-05, | |
| "loss": 0.0245, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.48645609617233276, | |
| "learning_rate": 5.167074885038373e-05, | |
| "loss": 0.024, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.1923076923076925, | |
| "grad_norm": 0.5603619813919067, | |
| "learning_rate": 5e-05, | |
| "loss": 0.0257, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.230769230769231, | |
| "grad_norm": 0.43165791034698486, | |
| "learning_rate": 4.832925114961629e-05, | |
| "loss": 0.0207, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.269230769230769, | |
| "grad_norm": 0.48327693343162537, | |
| "learning_rate": 4.666036831274392e-05, | |
| "loss": 0.0308, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.5035133957862854, | |
| "learning_rate": 4.4995215418795085e-05, | |
| "loss": 0.0333, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "eval_loss": 0.02246815524995327, | |
| "eval_runtime": 1.6386, | |
| "eval_samples_per_second": 13.426, | |
| "eval_steps_per_second": 1.221, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3461538461538463, | |
| "grad_norm": 0.38951441645622253, | |
| "learning_rate": 4.333565223131107e-05, | |
| "loss": 0.018, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.3846153846153846, | |
| "grad_norm": 0.5282712578773499, | |
| "learning_rate": 4.1683532270843504e-05, | |
| "loss": 0.0246, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.423076923076923, | |
| "grad_norm": 0.29735180735588074, | |
| "learning_rate": 4.0040700744808204e-05, | |
| "loss": 0.0165, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.3622055649757385, | |
| "learning_rate": 3.840899248662358e-05, | |
| "loss": 0.0203, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.39045998454093933, | |
| "learning_rate": 3.6790229906435705e-05, | |
| "loss": 0.0197, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "grad_norm": 0.39309993386268616, | |
| "learning_rate": 3.5186220955718306e-05, | |
| "loss": 0.0171, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5384615384615383, | |
| "eval_loss": 0.020282387733459473, | |
| "eval_runtime": 1.6665, | |
| "eval_samples_per_second": 13.202, | |
| "eval_steps_per_second": 1.2, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.5769230769230766, | |
| "grad_norm": 0.3418659269809723, | |
| "learning_rate": 3.3598757108021546e-05, | |
| "loss": 0.0112, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.4300234317779541, | |
| "learning_rate": 3.202961135812437e-05, | |
| "loss": 0.018, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.6538461538461537, | |
| "grad_norm": 0.34998640418052673, | |
| "learning_rate": 3.0480536241825263e-05, | |
| "loss": 0.014, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 0.46385401487350464, | |
| "learning_rate": 2.895326187858326e-05, | |
| "loss": 0.0178, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.730769230769231, | |
| "grad_norm": 0.4280989468097687, | |
| "learning_rate": 2.74494940391949e-05, | |
| "loss": 0.0195, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.337554395198822, | |
| "learning_rate": 2.5970912240665813e-05, | |
| "loss": 0.0172, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "eval_loss": 0.014381876215338707, | |
| "eval_runtime": 1.6934, | |
| "eval_samples_per_second": 12.991, | |
| "eval_steps_per_second": 1.181, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8076923076923075, | |
| "grad_norm": 0.32722559571266174, | |
| "learning_rate": 2.4519167870404125e-05, | |
| "loss": 0.0172, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.8461538461538463, | |
| "grad_norm": 0.3325071334838867, | |
| "learning_rate": 2.3095882341831372e-05, | |
| "loss": 0.0113, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.8846153846153846, | |
| "grad_norm": 0.31768810749053955, | |
| "learning_rate": 2.1702645283470236e-05, | |
| "loss": 0.0119, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.29802849888801575, | |
| "learning_rate": 2.0341012763532243e-05, | |
| "loss": 0.0131, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.9615384615384617, | |
| "grad_norm": 0.26436880230903625, | |
| "learning_rate": 1.9012505551987765e-05, | |
| "loss": 0.011, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.24293173849582672, | |
| "learning_rate": 1.771860742205988e-05, | |
| "loss": 0.0095, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.011883563362061977, | |
| "eval_runtime": 1.6658, | |
| "eval_samples_per_second": 13.207, | |
| "eval_steps_per_second": 1.201, | |
| "step": 78 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 104, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.039322719014093e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |