| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 16.93658536585366, | |
| "eval_steps": 500, | |
| "global_step": 136, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11707317073170732, | |
| "grad_norm": 1.7222386598587036, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 0.3206, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23414634146341465, | |
| "grad_norm": 1.8072729110717773, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.331, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.35121951219512193, | |
| "grad_norm": 1.7023462057113647, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.3156, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4682926829268293, | |
| "grad_norm": 1.6288352012634277, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.31, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5853658536585366, | |
| "grad_norm": 1.5489941835403442, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.312, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.7024390243902439, | |
| "grad_norm": 1.4443506002426147, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.2871, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.8195121951219512, | |
| "grad_norm": 1.3118906021118164, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3233, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.9365853658536586, | |
| "grad_norm": 1.110355257987976, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.2937, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1170731707317074, | |
| "grad_norm": 1.644830584526062, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 0.6554, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.2341463414634146, | |
| "grad_norm": 0.7390233874320984, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.2971, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.3512195121951218, | |
| "grad_norm": 0.49085134267807007, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 0.2825, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.4682926829268292, | |
| "grad_norm": 0.7516170144081116, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.2606, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.5853658536585367, | |
| "grad_norm": 0.8610337972640991, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 0.2717, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.7024390243902439, | |
| "grad_norm": 0.9654709696769714, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2672, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.819512195121951, | |
| "grad_norm": 0.8754584789276123, | |
| "learning_rate": 9.998342337571566e-06, | |
| "loss": 0.2758, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.9365853658536585, | |
| "grad_norm": 0.7926000356674194, | |
| "learning_rate": 9.993370449424153e-06, | |
| "loss": 0.2777, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.1170731707317074, | |
| "grad_norm": 1.2570271492004395, | |
| "learning_rate": 9.985087632242634e-06, | |
| "loss": 0.5213, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.234146341463415, | |
| "grad_norm": 0.47539860010147095, | |
| "learning_rate": 9.973499378072947e-06, | |
| "loss": 0.2423, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.351219512195122, | |
| "grad_norm": 0.3991779685020447, | |
| "learning_rate": 9.958613370680507e-06, | |
| "loss": 0.2236, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.4682926829268292, | |
| "grad_norm": 0.3700422942638397, | |
| "learning_rate": 9.940439480455386e-06, | |
| "loss": 0.2556, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.5853658536585367, | |
| "grad_norm": 0.3826175630092621, | |
| "learning_rate": 9.918989757867584e-06, | |
| "loss": 0.2441, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.7024390243902436, | |
| "grad_norm": 0.40204256772994995, | |
| "learning_rate": 9.89427842547679e-06, | |
| "loss": 0.2336, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.819512195121951, | |
| "grad_norm": 0.385053813457489, | |
| "learning_rate": 9.866321868501914e-06, | |
| "loss": 0.2435, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.9365853658536585, | |
| "grad_norm": 0.39567235112190247, | |
| "learning_rate": 9.835138623956603e-06, | |
| "loss": 0.2471, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.1170731707317074, | |
| "grad_norm": 0.6419874429702759, | |
| "learning_rate": 9.80074936835801e-06, | |
| "loss": 0.4381, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.234146341463415, | |
| "grad_norm": 0.2792779207229614, | |
| "learning_rate": 9.763176904016914e-06, | |
| "loss": 0.2104, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.351219512195122, | |
| "grad_norm": 0.2601935565471649, | |
| "learning_rate": 9.722446143918307e-06, | |
| "loss": 0.2092, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.4682926829268292, | |
| "grad_norm": 0.29861757159233093, | |
| "learning_rate": 9.678584095202468e-06, | |
| "loss": 0.1996, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5853658536585367, | |
| "grad_norm": 0.2728100121021271, | |
| "learning_rate": 9.631619841257477e-06, | |
| "loss": 0.1984, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.7024390243902436, | |
| "grad_norm": 0.23323267698287964, | |
| "learning_rate": 9.581584522435025e-06, | |
| "loss": 0.1921, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.819512195121951, | |
| "grad_norm": 0.2583334445953369, | |
| "learning_rate": 9.528511315402358e-06, | |
| "loss": 0.2008, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.9365853658536585, | |
| "grad_norm": 0.24806950986385345, | |
| "learning_rate": 9.472435411143979e-06, | |
| "loss": 0.2027, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.117073170731707, | |
| "grad_norm": 0.5137495994567871, | |
| "learning_rate": 9.413393991627737e-06, | |
| "loss": 0.3914, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 4.234146341463415, | |
| "grad_norm": 0.21133840084075928, | |
| "learning_rate": 9.351426205150778e-06, | |
| "loss": 0.1771, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.351219512195122, | |
| "grad_norm": 0.19555042684078217, | |
| "learning_rate": 9.286573140381663e-06, | |
| "loss": 0.1778, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.46829268292683, | |
| "grad_norm": 0.22236905992031097, | |
| "learning_rate": 9.218877799115929e-06, | |
| "loss": 0.1761, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.585365853658536, | |
| "grad_norm": 0.21382497251033783, | |
| "learning_rate": 9.148385067763094e-06, | |
| "loss": 0.1687, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 4.702439024390244, | |
| "grad_norm": 0.19119715690612793, | |
| "learning_rate": 9.075141687584056e-06, | |
| "loss": 0.159, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.819512195121951, | |
| "grad_norm": 0.23834119737148285, | |
| "learning_rate": 8.999196223698599e-06, | |
| "loss": 0.1538, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.9365853658536585, | |
| "grad_norm": 0.19232302904129028, | |
| "learning_rate": 8.920599032883553e-06, | |
| "loss": 0.1491, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.117073170731707, | |
| "grad_norm": 0.38959407806396484, | |
| "learning_rate": 8.839402230183e-06, | |
| "loss": 0.2932, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 5.234146341463415, | |
| "grad_norm": 0.17205432057380676, | |
| "learning_rate": 8.755659654352599e-06, | |
| "loss": 0.1403, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.351219512195122, | |
| "grad_norm": 0.16311697661876678, | |
| "learning_rate": 8.669426832160997e-06, | |
| "loss": 0.1409, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 5.46829268292683, | |
| "grad_norm": 0.1873118132352829, | |
| "learning_rate": 8.580760941571968e-06, | |
| "loss": 0.1271, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.585365853658536, | |
| "grad_norm": 0.18881656229496002, | |
| "learning_rate": 8.489720773831717e-06, | |
| "loss": 0.1233, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.702439024390244, | |
| "grad_norm": 0.17520631849765778, | |
| "learning_rate": 8.396366694486466e-06, | |
| "loss": 0.1275, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.819512195121951, | |
| "grad_norm": 0.15314437448978424, | |
| "learning_rate": 8.30076060335616e-06, | |
| "loss": 0.118, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.9365853658536585, | |
| "grad_norm": 0.15604296326637268, | |
| "learning_rate": 8.202965893490877e-06, | |
| "loss": 0.1163, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 6.117073170731707, | |
| "grad_norm": 0.17413058876991272, | |
| "learning_rate": 8.103047409137114e-06, | |
| "loss": 0.2231, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 6.234146341463415, | |
| "grad_norm": 0.16367758810520172, | |
| "learning_rate": 8.001071402741843e-06, | |
| "loss": 0.0976, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.351219512195122, | |
| "grad_norm": 0.15195311605930328, | |
| "learning_rate": 7.897105491022819e-06, | |
| "loss": 0.1004, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.46829268292683, | |
| "grad_norm": 0.18309926986694336, | |
| "learning_rate": 7.791218610134324e-06, | |
| "loss": 0.101, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.585365853658536, | |
| "grad_norm": 0.18088921904563904, | |
| "learning_rate": 7.683480969958005e-06, | |
| "loss": 0.0986, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 6.702439024390244, | |
| "grad_norm": 0.1502581387758255, | |
| "learning_rate": 7.5739640075491546e-06, | |
| "loss": 0.0866, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.819512195121951, | |
| "grad_norm": 0.14805804193019867, | |
| "learning_rate": 7.462740339769323e-06, | |
| "loss": 0.0839, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.9365853658536585, | |
| "grad_norm": 0.14805378019809723, | |
| "learning_rate": 7.349883715136601e-06, | |
| "loss": 0.0835, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 7.117073170731707, | |
| "grad_norm": 0.35682159662246704, | |
| "learning_rate": 7.235468964925571e-06, | |
| "loss": 0.1683, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 7.234146341463415, | |
| "grad_norm": 0.19683778285980225, | |
| "learning_rate": 7.119571953549305e-06, | |
| "loss": 0.0748, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 7.351219512195122, | |
| "grad_norm": 0.15031643211841583, | |
| "learning_rate": 7.002269528256334e-06, | |
| "loss": 0.0696, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.46829268292683, | |
| "grad_norm": 0.15439893305301666, | |
| "learning_rate": 6.883639468175926e-06, | |
| "loss": 0.0644, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.585365853658536, | |
| "grad_norm": 0.17578311264514923, | |
| "learning_rate": 6.763760432745475e-06, | |
| "loss": 0.0708, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 7.702439024390244, | |
| "grad_norm": 0.2290361076593399, | |
| "learning_rate": 6.6427119095541745e-06, | |
| "loss": 0.0685, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.819512195121951, | |
| "grad_norm": 0.12619617581367493, | |
| "learning_rate": 6.520574161637591e-06, | |
| "loss": 0.0685, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 7.9365853658536585, | |
| "grad_norm": 0.14427310228347778, | |
| "learning_rate": 6.397428174258048e-06, | |
| "loss": 0.0556, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 8.117073170731707, | |
| "grad_norm": 0.37869665026664734, | |
| "learning_rate": 6.273355601206143e-06, | |
| "loss": 0.0984, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 8.234146341463415, | |
| "grad_norm": 0.2276754379272461, | |
| "learning_rate": 6.148438710658979e-06, | |
| "loss": 0.0497, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.351219512195122, | |
| "grad_norm": 0.1407787799835205, | |
| "learning_rate": 6.022760330631006e-06, | |
| "loss": 0.0517, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 8.46829268292683, | |
| "grad_norm": 0.17464971542358398, | |
| "learning_rate": 5.896403794053679e-06, | |
| "loss": 0.0492, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.585365853658537, | |
| "grad_norm": 0.18100804090499878, | |
| "learning_rate": 5.76945288352031e-06, | |
| "loss": 0.0397, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 8.702439024390245, | |
| "grad_norm": 0.17883113026618958, | |
| "learning_rate": 5.641991775732756e-06, | |
| "loss": 0.0406, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.819512195121952, | |
| "grad_norm": 0.13799619674682617, | |
| "learning_rate": 5.514104985686802e-06, | |
| "loss": 0.0444, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 8.93658536585366, | |
| "grad_norm": 0.12878285348415375, | |
| "learning_rate": 5.385877310633233e-06, | |
| "loss": 0.0433, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 9.117073170731707, | |
| "grad_norm": 0.319606751203537, | |
| "learning_rate": 5.257393773851733e-06, | |
| "loss": 0.0843, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 9.234146341463415, | |
| "grad_norm": 0.1562907099723816, | |
| "learning_rate": 5.1287395682749444e-06, | |
| "loss": 0.0286, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 9.351219512195122, | |
| "grad_norm": 0.12641151249408722, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0249, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 9.46829268292683, | |
| "grad_norm": 0.11516842246055603, | |
| "learning_rate": 4.871260431725058e-06, | |
| "loss": 0.0359, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.585365853658537, | |
| "grad_norm": 0.1351480334997177, | |
| "learning_rate": 4.742606226148268e-06, | |
| "loss": 0.0413, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 9.702439024390245, | |
| "grad_norm": 0.1398896723985672, | |
| "learning_rate": 4.614122689366769e-06, | |
| "loss": 0.0354, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.819512195121952, | |
| "grad_norm": 0.13422027230262756, | |
| "learning_rate": 4.485895014313198e-06, | |
| "loss": 0.0273, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 9.93658536585366, | |
| "grad_norm": 0.10346952080726624, | |
| "learning_rate": 4.358008224267245e-06, | |
| "loss": 0.0362, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 10.117073170731707, | |
| "grad_norm": 0.22095316648483276, | |
| "learning_rate": 4.230547116479691e-06, | |
| "loss": 0.0408, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 10.234146341463415, | |
| "grad_norm": 0.08995038270950317, | |
| "learning_rate": 4.103596205946323e-06, | |
| "loss": 0.0265, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 10.351219512195122, | |
| "grad_norm": 0.10617168992757797, | |
| "learning_rate": 3.977239669368998e-06, | |
| "loss": 0.0217, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 10.46829268292683, | |
| "grad_norm": 0.09259112179279327, | |
| "learning_rate": 3.851561289341023e-06, | |
| "loss": 0.0237, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.585365853658537, | |
| "grad_norm": 0.09134888648986816, | |
| "learning_rate": 3.726644398793857e-06, | |
| "loss": 0.0205, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.702439024390245, | |
| "grad_norm": 0.08447619527578354, | |
| "learning_rate": 3.6025718257419532e-06, | |
| "loss": 0.0253, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.819512195121952, | |
| "grad_norm": 0.09264950454235077, | |
| "learning_rate": 3.4794258383624115e-06, | |
| "loss": 0.0254, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 10.93658536585366, | |
| "grad_norm": 0.10039179772138596, | |
| "learning_rate": 3.3572880904458267e-06, | |
| "loss": 0.0276, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 11.117073170731707, | |
| "grad_norm": 0.18901416659355164, | |
| "learning_rate": 3.236239567254526e-06, | |
| "loss": 0.0378, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 11.234146341463415, | |
| "grad_norm": 0.06545038521289825, | |
| "learning_rate": 3.116360531824074e-06, | |
| "loss": 0.0173, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 11.351219512195122, | |
| "grad_norm": 0.064559705555439, | |
| "learning_rate": 2.997730471743667e-06, | |
| "loss": 0.0165, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 11.46829268292683, | |
| "grad_norm": 0.07374051958322525, | |
| "learning_rate": 2.880428046450697e-06, | |
| "loss": 0.0191, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 11.585365853658537, | |
| "grad_norm": 0.06621921807527542, | |
| "learning_rate": 2.7645310350744296e-06, | |
| "loss": 0.0229, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 11.702439024390245, | |
| "grad_norm": 0.08165808767080307, | |
| "learning_rate": 2.6501162848634023e-06, | |
| "loss": 0.0202, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 11.819512195121952, | |
| "grad_norm": 0.09986157715320587, | |
| "learning_rate": 2.537259660230679e-06, | |
| "loss": 0.0192, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 11.93658536585366, | |
| "grad_norm": 0.07528764754533768, | |
| "learning_rate": 2.426035992450848e-06, | |
| "loss": 0.016, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 12.117073170731707, | |
| "grad_norm": 0.07593318819999695, | |
| "learning_rate": 2.316519030041998e-06, | |
| "loss": 0.0248, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 12.234146341463415, | |
| "grad_norm": 0.08455588668584824, | |
| "learning_rate": 2.2087813898656775e-06, | |
| "loss": 0.0143, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 12.351219512195122, | |
| "grad_norm": 0.07890263944864273, | |
| "learning_rate": 2.102894508977182e-06, | |
| "loss": 0.0149, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 12.46829268292683, | |
| "grad_norm": 0.05967545881867409, | |
| "learning_rate": 1.9989285972581595e-06, | |
| "loss": 0.0124, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.585365853658537, | |
| "grad_norm": 0.06302303820848465, | |
| "learning_rate": 1.896952590862886e-06, | |
| "loss": 0.0176, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 12.702439024390245, | |
| "grad_norm": 0.049937959760427475, | |
| "learning_rate": 1.7970341065091246e-06, | |
| "loss": 0.0114, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.819512195121952, | |
| "grad_norm": 0.05920058861374855, | |
| "learning_rate": 1.699239396643841e-06, | |
| "loss": 0.012, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 12.93658536585366, | |
| "grad_norm": 0.06141705811023712, | |
| "learning_rate": 1.6036333055135345e-06, | |
| "loss": 0.0156, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 13.117073170731707, | |
| "grad_norm": 0.20311079919338226, | |
| "learning_rate": 1.5102792261682813e-06, | |
| "loss": 0.0429, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 13.234146341463415, | |
| "grad_norm": 0.05362841859459877, | |
| "learning_rate": 1.4192390584280347e-06, | |
| "loss": 0.0106, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 13.351219512195122, | |
| "grad_norm": 0.04556548222899437, | |
| "learning_rate": 1.330573167839005e-06, | |
| "loss": 0.0141, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 13.46829268292683, | |
| "grad_norm": 0.0745721235871315, | |
| "learning_rate": 1.2443403456474017e-06, | |
| "loss": 0.0128, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 13.585365853658537, | |
| "grad_norm": 0.05052574723958969, | |
| "learning_rate": 1.1605977698170001e-06, | |
| "loss": 0.0118, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 13.702439024390245, | |
| "grad_norm": 0.06299788504838943, | |
| "learning_rate": 1.0794009671164484e-06, | |
| "loss": 0.0154, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 13.819512195121952, | |
| "grad_norm": 0.05452995002269745, | |
| "learning_rate": 1.0008037763014033e-06, | |
| "loss": 0.0097, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 13.93658536585366, | |
| "grad_norm": 0.05043210834264755, | |
| "learning_rate": 9.248583124159438e-07, | |
| "loss": 0.0111, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 14.117073170731707, | |
| "grad_norm": 0.10536748170852661, | |
| "learning_rate": 8.516149322369055e-07, | |
| "loss": 0.0138, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 14.234146341463415, | |
| "grad_norm": 0.04710059612989426, | |
| "learning_rate": 7.811222008840719e-07, | |
| "loss": 0.0113, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 14.351219512195122, | |
| "grad_norm": 0.04207915812730789, | |
| "learning_rate": 7.13426859618338e-07, | |
| "loss": 0.0114, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 14.46829268292683, | |
| "grad_norm": 0.0388207845389843, | |
| "learning_rate": 6.485737948492237e-07, | |
| "loss": 0.012, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 14.585365853658537, | |
| "grad_norm": 0.05471482500433922, | |
| "learning_rate": 5.866060083722624e-07, | |
| "loss": 0.0123, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 14.702439024390245, | |
| "grad_norm": 0.03734861686825752, | |
| "learning_rate": 5.275645888560233e-07, | |
| "loss": 0.0111, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 14.819512195121952, | |
| "grad_norm": 0.04156919941306114, | |
| "learning_rate": 4.71488684597643e-07, | |
| "loss": 0.0082, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 14.93658536585366, | |
| "grad_norm": 0.04900944232940674, | |
| "learning_rate": 4.184154775649768e-07, | |
| "loss": 0.0116, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 15.117073170731707, | |
| "grad_norm": 0.12125589698553085, | |
| "learning_rate": 3.683801587425251e-07, | |
| "loss": 0.0258, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 15.234146341463415, | |
| "grad_norm": 0.0401557981967926, | |
| "learning_rate": 3.214159047975324e-07, | |
| "loss": 0.0138, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 15.351219512195122, | |
| "grad_norm": 0.045213762670755386, | |
| "learning_rate": 2.7755385608169374e-07, | |
| "loss": 0.0108, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 15.46829268292683, | |
| "grad_norm": 0.03460492938756943, | |
| "learning_rate": 2.368230959830875e-07, | |
| "loss": 0.0074, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 15.585365853658537, | |
| "grad_norm": 0.035707153379917145, | |
| "learning_rate": 1.992506316419912e-07, | |
| "loss": 0.0096, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 15.702439024390245, | |
| "grad_norm": 0.037332240492105484, | |
| "learning_rate": 1.6486137604339813e-07, | |
| "loss": 0.0085, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 15.819512195121952, | |
| "grad_norm": 0.04060058668255806, | |
| "learning_rate": 1.3367813149808728e-07, | |
| "loss": 0.011, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 15.93658536585366, | |
| "grad_norm": 0.04075002670288086, | |
| "learning_rate": 1.0572157452321097e-07, | |
| "loss": 0.0105, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 16.117073170731707, | |
| "grad_norm": 0.11942698061466217, | |
| "learning_rate": 8.101024213241826e-08, | |
| "loss": 0.0211, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 16.234146341463415, | |
| "grad_norm": 0.03721379116177559, | |
| "learning_rate": 5.9560519544614725e-08, | |
| "loss": 0.0076, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 16.351219512195122, | |
| "grad_norm": 0.03629659488797188, | |
| "learning_rate": 4.138662931949255e-08, | |
| "loss": 0.0099, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 16.46829268292683, | |
| "grad_norm": 0.039185598492622375, | |
| "learning_rate": 2.6500621927054716e-08, | |
| "loss": 0.0094, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 16.585365853658537, | |
| "grad_norm": 0.04056056588888168, | |
| "learning_rate": 1.4912367757366485e-08, | |
| "loss": 0.0111, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 16.702439024390245, | |
| "grad_norm": 0.04743169620633125, | |
| "learning_rate": 6.629550575847355e-09, | |
| "loss": 0.0119, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 16.819512195121952, | |
| "grad_norm": 0.04899853095412254, | |
| "learning_rate": 1.657662428434792e-09, | |
| "loss": 0.0113, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 16.93658536585366, | |
| "grad_norm": 0.03317403793334961, | |
| "learning_rate": 0.0, | |
| "loss": 0.0115, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 16.93658536585366, | |
| "step": 136, | |
| "total_flos": 551725539786752.0, | |
| "train_loss": 0.11062119404703159, | |
| "train_runtime": 33672.3033, | |
| "train_samples_per_second": 0.414, | |
| "train_steps_per_second": 0.004 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 136, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 17, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 551725539786752.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |