| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 98, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02040816326530612, | |
| "grad_norm": 32.64888547751775, | |
| "learning_rate": 4e-07, | |
| "loss": 1.7523, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04081632653061224, | |
| "grad_norm": 29.727934420700674, | |
| "learning_rate": 8e-07, | |
| "loss": 1.7913, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.061224489795918366, | |
| "grad_norm": 24.405865809933943, | |
| "learning_rate": 1.2e-06, | |
| "loss": 1.7695, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.08163265306122448, | |
| "grad_norm": 10.393646720510844, | |
| "learning_rate": 1.6e-06, | |
| "loss": 1.7459, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 8.901802813033639, | |
| "learning_rate": 2e-06, | |
| "loss": 1.6905, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.12244897959183673, | |
| "grad_norm": 6.824537248098065, | |
| "learning_rate": 2.4e-06, | |
| "loss": 1.6515, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 7.83712617181299, | |
| "learning_rate": 2.8e-06, | |
| "loss": 1.6029, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.16326530612244897, | |
| "grad_norm": 6.032348523836469, | |
| "learning_rate": 3.2e-06, | |
| "loss": 1.6653, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1836734693877551, | |
| "grad_norm": 6.362729786136031, | |
| "learning_rate": 3.6e-06, | |
| "loss": 1.6055, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 5.363581180186887, | |
| "learning_rate": 4e-06, | |
| "loss": 1.626, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22448979591836735, | |
| "grad_norm": 5.614245135697976, | |
| "learning_rate": 3.99971472511942e-06, | |
| "loss": 1.6121, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.24489795918367346, | |
| "grad_norm": 5.004151493115574, | |
| "learning_rate": 3.998858981859435e-06, | |
| "loss": 1.6153, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2653061224489796, | |
| "grad_norm": 5.286918190889841, | |
| "learning_rate": 3.997433014342105e-06, | |
| "loss": 1.5694, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 4.875583241251316, | |
| "learning_rate": 3.995437229360142e-06, | |
| "loss": 1.6542, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 6.001720206242898, | |
| "learning_rate": 3.992872196260866e-06, | |
| "loss": 1.5836, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.32653061224489793, | |
| "grad_norm": 5.523386731381072, | |
| "learning_rate": 3.98973864678379e-06, | |
| "loss": 1.6299, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3469387755102041, | |
| "grad_norm": 9.219665630268889, | |
| "learning_rate": 3.986037474851867e-06, | |
| "loss": 1.6475, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3673469387755102, | |
| "grad_norm": 4.709359040696749, | |
| "learning_rate": 3.981769736316478e-06, | |
| "loss": 1.6115, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3877551020408163, | |
| "grad_norm": 4.977891858877406, | |
| "learning_rate": 3.976936648656223e-06, | |
| "loss": 1.5892, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 4.825081766216579, | |
| "learning_rate": 3.971539590629607e-06, | |
| "loss": 1.6161, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 4.521060581667052, | |
| "learning_rate": 3.965580101881716e-06, | |
| "loss": 1.5968, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.4489795918367347, | |
| "grad_norm": 5.559962293768277, | |
| "learning_rate": 3.959059882504989e-06, | |
| "loss": 1.574, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.46938775510204084, | |
| "grad_norm": 22.09612765494523, | |
| "learning_rate": 3.951980792554231e-06, | |
| "loss": 1.5748, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4897959183673469, | |
| "grad_norm": 5.40962977240307, | |
| "learning_rate": 3.944344851515981e-06, | |
| "loss": 1.6046, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 5.1355184641109455, | |
| "learning_rate": 3.936154237732409e-06, | |
| "loss": 1.621, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5306122448979592, | |
| "grad_norm": 6.8504360271944575, | |
| "learning_rate": 3.927411287779882e-06, | |
| "loss": 1.6124, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5510204081632653, | |
| "grad_norm": 4.482668388299841, | |
| "learning_rate": 3.918118495802404e-06, | |
| "loss": 1.5747, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 5.239757336297107, | |
| "learning_rate": 3.9082785128000975e-06, | |
| "loss": 1.5625, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5918367346938775, | |
| "grad_norm": 4.5776640976135585, | |
| "learning_rate": 3.8978941458729376e-06, | |
| "loss": 1.6, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 4.55153687355674, | |
| "learning_rate": 3.88696835741996e-06, | |
| "loss": 1.6203, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6326530612244898, | |
| "grad_norm": 4.489870280563155, | |
| "learning_rate": 3.875504264294161e-06, | |
| "loss": 1.6375, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6530612244897959, | |
| "grad_norm": 5.310641662353565, | |
| "learning_rate": 3.8635051369133365e-06, | |
| "loss": 1.5707, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.673469387755102, | |
| "grad_norm": 6.202748915782672, | |
| "learning_rate": 3.850974398327119e-06, | |
| "loss": 1.564, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6938775510204082, | |
| "grad_norm": 7.749515519920348, | |
| "learning_rate": 3.837915623240461e-06, | |
| "loss": 1.5765, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 5.415603692442806, | |
| "learning_rate": 3.824332536993866e-06, | |
| "loss": 1.5994, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7346938775510204, | |
| "grad_norm": 5.480596530632035, | |
| "learning_rate": 3.810229014500642e-06, | |
| "loss": 1.5561, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7551020408163265, | |
| "grad_norm": 4.233153242655977, | |
| "learning_rate": 3.795609079141483e-06, | |
| "loss": 1.5983, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.7755102040816326, | |
| "grad_norm": 4.178343085078006, | |
| "learning_rate": 3.780476901616703e-06, | |
| "loss": 1.6171, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7959183673469388, | |
| "grad_norm": 4.4176408442610455, | |
| "learning_rate": 3.764836798756438e-06, | |
| "loss": 1.5727, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 4.318784572510763, | |
| "learning_rate": 3.748693232289164e-06, | |
| "loss": 1.6577, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8367346938775511, | |
| "grad_norm": 4.683545812852939, | |
| "learning_rate": 3.7320508075688773e-06, | |
| "loss": 1.5839, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 4.2416142714136225, | |
| "learning_rate": 3.7149142722613015e-06, | |
| "loss": 1.5326, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.8775510204081632, | |
| "grad_norm": 5.8485689528715845, | |
| "learning_rate": 3.697288514989502e-06, | |
| "loss": 1.5531, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.8979591836734694, | |
| "grad_norm": 4.264324101534887, | |
| "learning_rate": 3.6791785639392775e-06, | |
| "loss": 1.6034, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 4.066575551440298, | |
| "learning_rate": 3.6605895854247527e-06, | |
| "loss": 1.5921, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.9387755102040817, | |
| "grad_norm": 4.034643806151124, | |
| "learning_rate": 3.641526882414553e-06, | |
| "loss": 1.5434, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.9591836734693877, | |
| "grad_norm": 4.132555724507058, | |
| "learning_rate": 3.6219958930190024e-06, | |
| "loss": 1.5853, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.9795918367346939, | |
| "grad_norm": 4.153531674893339, | |
| "learning_rate": 3.602002188938769e-06, | |
| "loss": 1.5858, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.146826988211348, | |
| "learning_rate": 3.581551473875397e-06, | |
| "loss": 1.5848, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.0204081632653061, | |
| "grad_norm": 8.656902757205804, | |
| "learning_rate": 3.5606495819041836e-06, | |
| "loss": 1.1151, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.0408163265306123, | |
| "grad_norm": 5.988124543792571, | |
| "learning_rate": 3.539302475809864e-06, | |
| "loss": 1.0798, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.0612244897959184, | |
| "grad_norm": 8.934677443634214, | |
| "learning_rate": 3.5175162453855814e-06, | |
| "loss": 1.043, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.0816326530612246, | |
| "grad_norm": 6.588611053883156, | |
| "learning_rate": 3.4952971056956184e-06, | |
| "loss": 1.0923, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.1020408163265305, | |
| "grad_norm": 5.749985643136262, | |
| "learning_rate": 3.4726513953023944e-06, | |
| "loss": 1.088, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.1224489795918366, | |
| "grad_norm": 5.539993882620073, | |
| "learning_rate": 3.4495855744582397e-06, | |
| "loss": 1.0816, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 6.90795649908795, | |
| "learning_rate": 3.4261062232624404e-06, | |
| "loss": 0.9693, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.163265306122449, | |
| "grad_norm": 5.478500360298918, | |
| "learning_rate": 3.4022200397841054e-06, | |
| "loss": 1.015, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.183673469387755, | |
| "grad_norm": 6.342399558155308, | |
| "learning_rate": 3.3779338381513733e-06, | |
| "loss": 0.992, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.2040816326530612, | |
| "grad_norm": 5.701765352521379, | |
| "learning_rate": 3.3532545466075147e-06, | |
| "loss": 0.9691, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.2244897959183674, | |
| "grad_norm": 5.5584312567296035, | |
| "learning_rate": 3.328189205534478e-06, | |
| "loss": 0.9681, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.2448979591836735, | |
| "grad_norm": 7.038510873972207, | |
| "learning_rate": 3.3027449654444447e-06, | |
| "loss": 0.9677, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.2653061224489797, | |
| "grad_norm": 5.392994856020097, | |
| "learning_rate": 3.276929084939967e-06, | |
| "loss": 0.9885, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 5.568535449393245, | |
| "learning_rate": 3.2507489286432733e-06, | |
| "loss": 0.993, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.306122448979592, | |
| "grad_norm": 5.1121872396505745, | |
| "learning_rate": 3.2242119650953256e-06, | |
| "loss": 0.9274, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.3265306122448979, | |
| "grad_norm": 5.038687234249487, | |
| "learning_rate": 3.19732576462523e-06, | |
| "loss": 1.0066, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.346938775510204, | |
| "grad_norm": 5.686879366948617, | |
| "learning_rate": 3.170097997190615e-06, | |
| "loss": 1.0032, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.3673469387755102, | |
| "grad_norm": 5.312762602694415, | |
| "learning_rate": 3.1425364301895844e-06, | |
| "loss": 0.9553, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.3877551020408163, | |
| "grad_norm": 4.940862521459705, | |
| "learning_rate": 3.1146489262448726e-06, | |
| "loss": 0.9473, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.4081632653061225, | |
| "grad_norm": 5.1441112950497745, | |
| "learning_rate": 3.0864434409608376e-06, | |
| "loss": 1.0159, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 5.485127198500315, | |
| "learning_rate": 3.0579280206539247e-06, | |
| "loss": 0.9193, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.4489795918367347, | |
| "grad_norm": 5.438513772229916, | |
| "learning_rate": 3.029110800057258e-06, | |
| "loss": 1.0163, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.469387755102041, | |
| "grad_norm": 5.089093472934727, | |
| "learning_rate": 3e-06, | |
| "loss": 1.0361, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.489795918367347, | |
| "grad_norm": 4.971637129071151, | |
| "learning_rate": 2.970603925062162e-06, | |
| "loss": 0.973, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.510204081632653, | |
| "grad_norm": 5.336082215964221, | |
| "learning_rate": 2.9409309612055114e-06, | |
| "loss": 1.0018, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.5306122448979593, | |
| "grad_norm": 4.726728122616708, | |
| "learning_rate": 2.9109895733812677e-06, | |
| "loss": 0.9467, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.5510204081632653, | |
| "grad_norm": 5.658444364153749, | |
| "learning_rate": 2.8807883031152685e-06, | |
| "loss": 0.9402, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 5.318192175322547, | |
| "learning_rate": 2.850335766071281e-06, | |
| "loss": 0.9771, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.5918367346938775, | |
| "grad_norm": 5.15866054072964, | |
| "learning_rate": 2.8196406495931747e-06, | |
| "loss": 0.9311, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.6122448979591837, | |
| "grad_norm": 5.122060161147712, | |
| "learning_rate": 2.7887117102266372e-06, | |
| "loss": 0.9289, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.6326530612244898, | |
| "grad_norm": 5.365443682206108, | |
| "learning_rate": 2.757557771221152e-06, | |
| "loss": 1.0086, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.6530612244897958, | |
| "grad_norm": 5.578125995371937, | |
| "learning_rate": 2.726187720012949e-06, | |
| "loss": 1.0213, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.6734693877551021, | |
| "grad_norm": 4.867474459234814, | |
| "learning_rate": 2.69461050568964e-06, | |
| "loss": 0.9611, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.693877551020408, | |
| "grad_norm": 5.134575732180386, | |
| "learning_rate": 2.6628351364372716e-06, | |
| "loss": 0.9465, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 4.929124345246347, | |
| "learning_rate": 2.6308706769705113e-06, | |
| "loss": 0.9854, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.7346938775510203, | |
| "grad_norm": 5.361881799554932, | |
| "learning_rate": 2.5987262459467165e-06, | |
| "loss": 0.9219, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.7551020408163265, | |
| "grad_norm": 5.111629187737712, | |
| "learning_rate": 2.5664110133646076e-06, | |
| "loss": 0.9139, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.7755102040816326, | |
| "grad_norm": 5.233284173002549, | |
| "learning_rate": 2.533934197948303e-06, | |
| "loss": 0.9921, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.7959183673469388, | |
| "grad_norm": 4.986446843273751, | |
| "learning_rate": 2.501305064517441e-06, | |
| "loss": 0.9397, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.816326530612245, | |
| "grad_norm": 4.975196749685198, | |
| "learning_rate": 2.468532921344164e-06, | |
| "loss": 0.9643, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.836734693877551, | |
| "grad_norm": 5.0457233401836135, | |
| "learning_rate": 2.4356271174977026e-06, | |
| "loss": 0.9169, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 5.534188821264441, | |
| "learning_rate": 2.40259704017732e-06, | |
| "loss": 0.9826, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.8775510204081631, | |
| "grad_norm": 5.119128865484249, | |
| "learning_rate": 2.369452112034379e-06, | |
| "loss": 0.8978, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.8979591836734695, | |
| "grad_norm": 5.059558900584382, | |
| "learning_rate": 2.3362017884842965e-06, | |
| "loss": 0.9653, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.9183673469387754, | |
| "grad_norm": 5.418073717960386, | |
| "learning_rate": 2.3028555550091533e-06, | |
| "loss": 0.924, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.9387755102040818, | |
| "grad_norm": 5.0451248325321805, | |
| "learning_rate": 2.2694229244517222e-06, | |
| "loss": 1.0111, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.9591836734693877, | |
| "grad_norm": 5.143777913795045, | |
| "learning_rate": 2.2359134343016923e-06, | |
| "loss": 0.9129, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.9795918367346939, | |
| "grad_norm": 5.05591838424157, | |
| "learning_rate": 2.2023366439748643e-06, | |
| "loss": 0.9551, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.043968948032216, | |
| "learning_rate": 2.1687021320860892e-06, | |
| "loss": 0.9724, | |
| "step": 98 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 196, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 49, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 26753888157696.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |