| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.21606648199446, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0110803324099723, | |
| "grad_norm": 1.0439653396606445, | |
| "learning_rate": 0.0001, | |
| "loss": 2.5223, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0221606648199446, | |
| "grad_norm": 1.072291612625122, | |
| "learning_rate": 9.949748743718594e-05, | |
| "loss": 2.465, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0332409972299169, | |
| "grad_norm": 0.9897931218147278, | |
| "learning_rate": 9.899497487437186e-05, | |
| "loss": 2.4409, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0443213296398892, | |
| "grad_norm": 1.0319087505340576, | |
| "learning_rate": 9.84924623115578e-05, | |
| "loss": 2.3619, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.055401662049861494, | |
| "grad_norm": 1.0002905130386353, | |
| "learning_rate": 9.798994974874372e-05, | |
| "loss": 2.1627, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0664819944598338, | |
| "grad_norm": 1.0927557945251465, | |
| "learning_rate": 9.748743718592965e-05, | |
| "loss": 2.0505, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07756232686980609, | |
| "grad_norm": 0.9562727808952332, | |
| "learning_rate": 9.698492462311559e-05, | |
| "loss": 1.9197, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0886426592797784, | |
| "grad_norm": 1.1700633764266968, | |
| "learning_rate": 9.64824120603015e-05, | |
| "loss": 1.6867, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0997229916897507, | |
| "grad_norm": 0.8975285291671753, | |
| "learning_rate": 9.597989949748745e-05, | |
| "loss": 1.6386, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.11080332409972299, | |
| "grad_norm": 0.8447593450546265, | |
| "learning_rate": 9.547738693467337e-05, | |
| "loss": 1.522, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12188365650969529, | |
| "grad_norm": 0.8749620318412781, | |
| "learning_rate": 9.49748743718593e-05, | |
| "loss": 1.4721, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1329639889196676, | |
| "grad_norm": 0.6403835415840149, | |
| "learning_rate": 9.447236180904523e-05, | |
| "loss": 1.4496, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1440443213296399, | |
| "grad_norm": 0.4993043541908264, | |
| "learning_rate": 9.396984924623115e-05, | |
| "loss": 1.4022, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15512465373961218, | |
| "grad_norm": 0.4994354844093323, | |
| "learning_rate": 9.34673366834171e-05, | |
| "loss": 1.4015, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16620498614958448, | |
| "grad_norm": 0.5358327627182007, | |
| "learning_rate": 9.296482412060302e-05, | |
| "loss": 1.2465, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1772853185595568, | |
| "grad_norm": 0.5239339470863342, | |
| "learning_rate": 9.246231155778895e-05, | |
| "loss": 1.3862, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1883656509695291, | |
| "grad_norm": 0.5605911612510681, | |
| "learning_rate": 9.195979899497488e-05, | |
| "loss": 1.3572, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1994459833795014, | |
| "grad_norm": 0.5271956324577332, | |
| "learning_rate": 9.14572864321608e-05, | |
| "loss": 1.363, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 0.5072754621505737, | |
| "learning_rate": 9.095477386934675e-05, | |
| "loss": 1.3393, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.22160664819944598, | |
| "grad_norm": 0.5984258055686951, | |
| "learning_rate": 9.045226130653267e-05, | |
| "loss": 1.4338, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23268698060941828, | |
| "grad_norm": 0.6067867279052734, | |
| "learning_rate": 8.99497487437186e-05, | |
| "loss": 1.1891, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.24376731301939059, | |
| "grad_norm": 0.5885209441184998, | |
| "learning_rate": 8.944723618090453e-05, | |
| "loss": 1.3141, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2548476454293629, | |
| "grad_norm": 0.5784013271331787, | |
| "learning_rate": 8.894472361809045e-05, | |
| "loss": 1.4002, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2659279778393352, | |
| "grad_norm": 0.6238617897033691, | |
| "learning_rate": 8.84422110552764e-05, | |
| "loss": 1.3739, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2770083102493075, | |
| "grad_norm": 0.6549237370491028, | |
| "learning_rate": 8.793969849246232e-05, | |
| "loss": 1.4349, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2880886426592798, | |
| "grad_norm": 0.6756062507629395, | |
| "learning_rate": 8.743718592964825e-05, | |
| "loss": 1.3734, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.29916897506925205, | |
| "grad_norm": 0.7228646278381348, | |
| "learning_rate": 8.693467336683418e-05, | |
| "loss": 1.3946, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.31024930747922436, | |
| "grad_norm": 0.6804022192955017, | |
| "learning_rate": 8.64321608040201e-05, | |
| "loss": 1.3068, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.32132963988919666, | |
| "grad_norm": 0.7463417053222656, | |
| "learning_rate": 8.592964824120603e-05, | |
| "loss": 1.3368, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.33240997229916897, | |
| "grad_norm": 0.7411599159240723, | |
| "learning_rate": 8.542713567839196e-05, | |
| "loss": 1.3373, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.34349030470914127, | |
| "grad_norm": 0.7613984942436218, | |
| "learning_rate": 8.49246231155779e-05, | |
| "loss": 1.2158, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3545706371191136, | |
| "grad_norm": 0.8660432696342468, | |
| "learning_rate": 8.442211055276383e-05, | |
| "loss": 1.2104, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3656509695290859, | |
| "grad_norm": 0.8703951835632324, | |
| "learning_rate": 8.391959798994975e-05, | |
| "loss": 1.2992, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3767313019390582, | |
| "grad_norm": 0.8799692988395691, | |
| "learning_rate": 8.341708542713568e-05, | |
| "loss": 1.2525, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3878116343490305, | |
| "grad_norm": 0.8259297609329224, | |
| "learning_rate": 8.291457286432161e-05, | |
| "loss": 1.2281, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3988919667590028, | |
| "grad_norm": 0.9279691576957703, | |
| "learning_rate": 8.241206030150754e-05, | |
| "loss": 1.3589, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4099722991689751, | |
| "grad_norm": 0.8066436648368835, | |
| "learning_rate": 8.190954773869348e-05, | |
| "loss": 1.2785, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.7177894115447998, | |
| "learning_rate": 8.14070351758794e-05, | |
| "loss": 1.278, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.43213296398891965, | |
| "grad_norm": 0.6018456220626831, | |
| "learning_rate": 8.090452261306533e-05, | |
| "loss": 1.2151, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.44321329639889195, | |
| "grad_norm": 0.5777440667152405, | |
| "learning_rate": 8.040201005025126e-05, | |
| "loss": 1.235, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.45429362880886426, | |
| "grad_norm": 0.4722194969654083, | |
| "learning_rate": 7.989949748743719e-05, | |
| "loss": 1.256, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.46537396121883656, | |
| "grad_norm": 0.4909801185131073, | |
| "learning_rate": 7.939698492462313e-05, | |
| "loss": 1.2826, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.47645429362880887, | |
| "grad_norm": 0.5409879684448242, | |
| "learning_rate": 7.889447236180904e-05, | |
| "loss": 1.2748, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.48753462603878117, | |
| "grad_norm": 0.5462285876274109, | |
| "learning_rate": 7.839195979899498e-05, | |
| "loss": 1.2736, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4986149584487535, | |
| "grad_norm": 0.5091783404350281, | |
| "learning_rate": 7.788944723618091e-05, | |
| "loss": 1.1858, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5096952908587258, | |
| "grad_norm": 0.4923122227191925, | |
| "learning_rate": 7.738693467336684e-05, | |
| "loss": 1.2042, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5207756232686981, | |
| "grad_norm": 0.5117873549461365, | |
| "learning_rate": 7.688442211055277e-05, | |
| "loss": 1.2803, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5318559556786704, | |
| "grad_norm": 0.47855451703071594, | |
| "learning_rate": 7.638190954773869e-05, | |
| "loss": 1.2136, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.5429362880886427, | |
| "grad_norm": 0.4981115460395813, | |
| "learning_rate": 7.587939698492463e-05, | |
| "loss": 1.2662, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.554016620498615, | |
| "grad_norm": 0.4743058681488037, | |
| "learning_rate": 7.537688442211056e-05, | |
| "loss": 1.1551, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5650969529085873, | |
| "grad_norm": 0.4888276159763336, | |
| "learning_rate": 7.487437185929649e-05, | |
| "loss": 1.1727, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5761772853185596, | |
| "grad_norm": 0.5036386251449585, | |
| "learning_rate": 7.437185929648241e-05, | |
| "loss": 1.2205, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5872576177285319, | |
| "grad_norm": 0.5121099352836609, | |
| "learning_rate": 7.386934673366834e-05, | |
| "loss": 1.2157, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5983379501385041, | |
| "grad_norm": 0.5122838616371155, | |
| "learning_rate": 7.336683417085427e-05, | |
| "loss": 1.2199, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.6094182825484764, | |
| "grad_norm": 0.5164381265640259, | |
| "learning_rate": 7.28643216080402e-05, | |
| "loss": 1.2962, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6204986149584487, | |
| "grad_norm": 0.5261529684066772, | |
| "learning_rate": 7.236180904522614e-05, | |
| "loss": 1.2283, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.5181484818458557, | |
| "learning_rate": 7.185929648241206e-05, | |
| "loss": 1.2724, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6426592797783933, | |
| "grad_norm": 0.48683488368988037, | |
| "learning_rate": 7.135678391959799e-05, | |
| "loss": 1.3003, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6537396121883656, | |
| "grad_norm": 0.4646837115287781, | |
| "learning_rate": 7.085427135678392e-05, | |
| "loss": 1.2103, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6648199445983379, | |
| "grad_norm": 0.513752818107605, | |
| "learning_rate": 7.035175879396985e-05, | |
| "loss": 1.1334, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6759002770083102, | |
| "grad_norm": 0.47100791335105896, | |
| "learning_rate": 6.984924623115579e-05, | |
| "loss": 1.2148, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6869806094182825, | |
| "grad_norm": 0.5160151124000549, | |
| "learning_rate": 6.93467336683417e-05, | |
| "loss": 1.2718, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6980609418282548, | |
| "grad_norm": 0.42398396134376526, | |
| "learning_rate": 6.884422110552764e-05, | |
| "loss": 1.119, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.7091412742382271, | |
| "grad_norm": 0.5424822568893433, | |
| "learning_rate": 6.834170854271357e-05, | |
| "loss": 1.2911, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7202216066481995, | |
| "grad_norm": 0.557855486869812, | |
| "learning_rate": 6.78391959798995e-05, | |
| "loss": 1.1695, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7313019390581718, | |
| "grad_norm": 0.4864124059677124, | |
| "learning_rate": 6.733668341708544e-05, | |
| "loss": 1.1762, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7423822714681441, | |
| "grad_norm": 0.5386707186698914, | |
| "learning_rate": 6.683417085427135e-05, | |
| "loss": 1.1766, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7534626038781164, | |
| "grad_norm": 0.5114085674285889, | |
| "learning_rate": 6.633165829145729e-05, | |
| "loss": 1.2149, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7645429362880887, | |
| "grad_norm": 0.5140852928161621, | |
| "learning_rate": 6.582914572864322e-05, | |
| "loss": 1.2486, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.775623268698061, | |
| "grad_norm": 0.4860430359840393, | |
| "learning_rate": 6.532663316582915e-05, | |
| "loss": 1.2724, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7867036011080333, | |
| "grad_norm": 0.5633051991462708, | |
| "learning_rate": 6.482412060301508e-05, | |
| "loss": 1.2467, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7977839335180056, | |
| "grad_norm": 0.48254477977752686, | |
| "learning_rate": 6.4321608040201e-05, | |
| "loss": 1.1976, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.8088642659279779, | |
| "grad_norm": 0.48218846321105957, | |
| "learning_rate": 6.381909547738694e-05, | |
| "loss": 1.1644, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8199445983379502, | |
| "grad_norm": 0.4937390387058258, | |
| "learning_rate": 6.331658291457287e-05, | |
| "loss": 1.2049, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8310249307479224, | |
| "grad_norm": 0.48136794567108154, | |
| "learning_rate": 6.28140703517588e-05, | |
| "loss": 1.2073, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.5367251038551331, | |
| "learning_rate": 6.231155778894473e-05, | |
| "loss": 1.3024, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.853185595567867, | |
| "grad_norm": 0.5513749718666077, | |
| "learning_rate": 6.180904522613065e-05, | |
| "loss": 1.2312, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8642659279778393, | |
| "grad_norm": 0.5120927095413208, | |
| "learning_rate": 6.130653266331658e-05, | |
| "loss": 1.1814, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8753462603878116, | |
| "grad_norm": 0.47654178738594055, | |
| "learning_rate": 6.080402010050251e-05, | |
| "loss": 1.1599, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8864265927977839, | |
| "grad_norm": 0.5559302568435669, | |
| "learning_rate": 6.030150753768844e-05, | |
| "loss": 1.2203, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8975069252077562, | |
| "grad_norm": 0.5184886455535889, | |
| "learning_rate": 5.979899497487438e-05, | |
| "loss": 1.2238, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.9085872576177285, | |
| "grad_norm": 0.5314403176307678, | |
| "learning_rate": 5.929648241206031e-05, | |
| "loss": 1.2255, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.9196675900277008, | |
| "grad_norm": 0.4995604455471039, | |
| "learning_rate": 5.879396984924623e-05, | |
| "loss": 1.1613, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9307479224376731, | |
| "grad_norm": 0.483528733253479, | |
| "learning_rate": 5.829145728643216e-05, | |
| "loss": 1.2412, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9418282548476454, | |
| "grad_norm": 0.5035815238952637, | |
| "learning_rate": 5.778894472361809e-05, | |
| "loss": 1.2189, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9529085872576177, | |
| "grad_norm": 0.5537089705467224, | |
| "learning_rate": 5.728643216080403e-05, | |
| "loss": 1.3265, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.96398891966759, | |
| "grad_norm": 0.5601852536201477, | |
| "learning_rate": 5.6783919597989955e-05, | |
| "loss": 1.2444, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9750692520775623, | |
| "grad_norm": 0.5105345249176025, | |
| "learning_rate": 5.628140703517588e-05, | |
| "loss": 1.169, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9861495844875346, | |
| "grad_norm": 0.533295750617981, | |
| "learning_rate": 5.577889447236181e-05, | |
| "loss": 1.1606, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.997229916897507, | |
| "grad_norm": 0.5628755688667297, | |
| "learning_rate": 5.527638190954774e-05, | |
| "loss": 1.2022, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0083102493074791, | |
| "grad_norm": 1.1733123064041138, | |
| "learning_rate": 5.477386934673368e-05, | |
| "loss": 2.0244, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0193905817174516, | |
| "grad_norm": 0.48274314403533936, | |
| "learning_rate": 5.4271356783919604e-05, | |
| "loss": 1.1683, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0304709141274238, | |
| "grad_norm": 0.5582519173622131, | |
| "learning_rate": 5.376884422110553e-05, | |
| "loss": 1.3466, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0415512465373962, | |
| "grad_norm": 0.4637548327445984, | |
| "learning_rate": 5.3266331658291455e-05, | |
| "loss": 1.0639, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.5016134977340698, | |
| "learning_rate": 5.276381909547739e-05, | |
| "loss": 1.2031, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0637119113573408, | |
| "grad_norm": 0.5333296060562134, | |
| "learning_rate": 5.226130653266332e-05, | |
| "loss": 1.145, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.074792243767313, | |
| "grad_norm": 0.5505311489105225, | |
| "learning_rate": 5.175879396984925e-05, | |
| "loss": 1.2017, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.0858725761772854, | |
| "grad_norm": 0.49051010608673096, | |
| "learning_rate": 5.125628140703518e-05, | |
| "loss": 1.1612, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0969529085872576, | |
| "grad_norm": 0.526991605758667, | |
| "learning_rate": 5.0753768844221104e-05, | |
| "loss": 1.1927, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.10803324099723, | |
| "grad_norm": 0.5483986139297485, | |
| "learning_rate": 5.0251256281407036e-05, | |
| "loss": 1.2416, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1191135734072022, | |
| "grad_norm": 0.5175971984863281, | |
| "learning_rate": 4.974874371859297e-05, | |
| "loss": 1.097, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.1301939058171746, | |
| "grad_norm": 0.5794366002082825, | |
| "learning_rate": 4.92462311557789e-05, | |
| "loss": 1.1469, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.1412742382271468, | |
| "grad_norm": 0.5408708453178406, | |
| "learning_rate": 4.874371859296483e-05, | |
| "loss": 1.1193, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1523545706371192, | |
| "grad_norm": 0.504085123538971, | |
| "learning_rate": 4.824120603015075e-05, | |
| "loss": 0.9672, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1634349030470914, | |
| "grad_norm": 0.6266749501228333, | |
| "learning_rate": 4.7738693467336685e-05, | |
| "loss": 1.1375, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1745152354570636, | |
| "grad_norm": 0.5513699054718018, | |
| "learning_rate": 4.723618090452262e-05, | |
| "loss": 1.2081, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.185595567867036, | |
| "grad_norm": 0.5879850387573242, | |
| "learning_rate": 4.673366834170855e-05, | |
| "loss": 1.1576, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.1966759002770084, | |
| "grad_norm": 0.5555039048194885, | |
| "learning_rate": 4.6231155778894475e-05, | |
| "loss": 1.1212, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.2077562326869806, | |
| "grad_norm": 0.5815752744674683, | |
| "learning_rate": 4.57286432160804e-05, | |
| "loss": 1.1984, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.2188365650969528, | |
| "grad_norm": 0.6069645881652832, | |
| "learning_rate": 4.522613065326633e-05, | |
| "loss": 1.1879, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2299168975069252, | |
| "grad_norm": 0.6030775308609009, | |
| "learning_rate": 4.4723618090452266e-05, | |
| "loss": 1.1959, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2409972299168974, | |
| "grad_norm": 0.5615729093551636, | |
| "learning_rate": 4.42211055276382e-05, | |
| "loss": 1.1266, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2520775623268698, | |
| "grad_norm": 0.5482991337776184, | |
| "learning_rate": 4.3718592964824124e-05, | |
| "loss": 1.158, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.6517236232757568, | |
| "learning_rate": 4.321608040201005e-05, | |
| "loss": 1.2302, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2742382271468145, | |
| "grad_norm": 0.5991782546043396, | |
| "learning_rate": 4.271356783919598e-05, | |
| "loss": 1.1322, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2853185595567866, | |
| "grad_norm": 0.6047670841217041, | |
| "learning_rate": 4.2211055276381914e-05, | |
| "loss": 1.1633, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.296398891966759, | |
| "grad_norm": 0.6029966473579407, | |
| "learning_rate": 4.170854271356784e-05, | |
| "loss": 1.1269, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.3074792243767313, | |
| "grad_norm": 0.5472580790519714, | |
| "learning_rate": 4.120603015075377e-05, | |
| "loss": 1.0763, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.3185595567867037, | |
| "grad_norm": 0.5828231573104858, | |
| "learning_rate": 4.07035175879397e-05, | |
| "loss": 1.1432, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3296398891966759, | |
| "grad_norm": 0.6450657844543457, | |
| "learning_rate": 4.020100502512563e-05, | |
| "loss": 1.2425, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3407202216066483, | |
| "grad_norm": 0.6098791360855103, | |
| "learning_rate": 3.969849246231156e-05, | |
| "loss": 1.0802, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.3518005540166205, | |
| "grad_norm": 0.5955787897109985, | |
| "learning_rate": 3.919597989949749e-05, | |
| "loss": 1.0539, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3628808864265927, | |
| "grad_norm": 0.651118814945221, | |
| "learning_rate": 3.869346733668342e-05, | |
| "loss": 1.2524, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.373961218836565, | |
| "grad_norm": 0.6121578216552734, | |
| "learning_rate": 3.8190954773869346e-05, | |
| "loss": 1.1569, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3850415512465375, | |
| "grad_norm": 0.6717909574508667, | |
| "learning_rate": 3.768844221105528e-05, | |
| "loss": 1.2191, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3961218836565097, | |
| "grad_norm": 0.6332257390022278, | |
| "learning_rate": 3.7185929648241204e-05, | |
| "loss": 1.1298, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.4072022160664819, | |
| "grad_norm": 0.6334303617477417, | |
| "learning_rate": 3.668341708542714e-05, | |
| "loss": 1.0351, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.4182825484764543, | |
| "grad_norm": 0.655293345451355, | |
| "learning_rate": 3.618090452261307e-05, | |
| "loss": 1.177, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.4293628808864267, | |
| "grad_norm": 0.6217477321624756, | |
| "learning_rate": 3.5678391959798995e-05, | |
| "loss": 1.1263, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.440443213296399, | |
| "grad_norm": 0.6945567727088928, | |
| "learning_rate": 3.517587939698493e-05, | |
| "loss": 1.2184, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.451523545706371, | |
| "grad_norm": 0.6126496195793152, | |
| "learning_rate": 3.467336683417085e-05, | |
| "loss": 1.0179, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4626038781163435, | |
| "grad_norm": 0.6645523309707642, | |
| "learning_rate": 3.4170854271356785e-05, | |
| "loss": 1.1921, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.65655916929245, | |
| "learning_rate": 3.366834170854272e-05, | |
| "loss": 1.1358, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4847645429362881, | |
| "grad_norm": 0.6353817582130432, | |
| "learning_rate": 3.3165829145728643e-05, | |
| "loss": 1.0577, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4958448753462603, | |
| "grad_norm": 0.6068665385246277, | |
| "learning_rate": 3.2663316582914576e-05, | |
| "loss": 1.1409, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.5069252077562327, | |
| "grad_norm": 0.6636937856674194, | |
| "learning_rate": 3.21608040201005e-05, | |
| "loss": 1.1746, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.5180055401662051, | |
| "grad_norm": 0.764776885509491, | |
| "learning_rate": 3.1658291457286434e-05, | |
| "loss": 1.2315, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.5290858725761773, | |
| "grad_norm": 0.6621037125587463, | |
| "learning_rate": 3.1155778894472366e-05, | |
| "loss": 1.122, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5401662049861495, | |
| "grad_norm": 0.6436090469360352, | |
| "learning_rate": 3.065326633165829e-05, | |
| "loss": 1.0881, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5512465373961217, | |
| "grad_norm": 0.6181167364120483, | |
| "learning_rate": 3.015075376884422e-05, | |
| "loss": 1.1523, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5623268698060941, | |
| "grad_norm": 0.6687259674072266, | |
| "learning_rate": 2.9648241206030153e-05, | |
| "loss": 1.0796, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.5734072022160666, | |
| "grad_norm": 0.6394988298416138, | |
| "learning_rate": 2.914572864321608e-05, | |
| "loss": 1.1142, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.5844875346260388, | |
| "grad_norm": 0.6827041506767273, | |
| "learning_rate": 2.8643216080402015e-05, | |
| "loss": 1.1727, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.595567867036011, | |
| "grad_norm": 0.6124918460845947, | |
| "learning_rate": 2.814070351758794e-05, | |
| "loss": 1.037, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.6066481994459834, | |
| "grad_norm": 0.6157639622688293, | |
| "learning_rate": 2.763819095477387e-05, | |
| "loss": 1.0598, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.6177285318559558, | |
| "grad_norm": 0.667032778263092, | |
| "learning_rate": 2.7135678391959802e-05, | |
| "loss": 1.1222, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.628808864265928, | |
| "grad_norm": 0.7053226232528687, | |
| "learning_rate": 2.6633165829145728e-05, | |
| "loss": 1.2135, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6398891966759002, | |
| "grad_norm": 0.7048087120056152, | |
| "learning_rate": 2.613065326633166e-05, | |
| "loss": 1.1588, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6509695290858726, | |
| "grad_norm": 0.6474671959877014, | |
| "learning_rate": 2.562814070351759e-05, | |
| "loss": 1.0781, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.662049861495845, | |
| "grad_norm": 0.65389084815979, | |
| "learning_rate": 2.5125628140703518e-05, | |
| "loss": 1.0982, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6731301939058172, | |
| "grad_norm": 0.6817282438278198, | |
| "learning_rate": 2.462311557788945e-05, | |
| "loss": 1.0967, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.6355194449424744, | |
| "learning_rate": 2.4120603015075376e-05, | |
| "loss": 1.0279, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6952908587257618, | |
| "grad_norm": 0.683628261089325, | |
| "learning_rate": 2.361809045226131e-05, | |
| "loss": 1.1533, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.7063711911357342, | |
| "grad_norm": 0.710445761680603, | |
| "learning_rate": 2.3115577889447238e-05, | |
| "loss": 1.182, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.7174515235457064, | |
| "grad_norm": 0.6544257998466492, | |
| "learning_rate": 2.2613065326633167e-05, | |
| "loss": 1.0936, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7285318559556786, | |
| "grad_norm": 0.6538469791412354, | |
| "learning_rate": 2.21105527638191e-05, | |
| "loss": 1.0657, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.739612188365651, | |
| "grad_norm": 0.6852319836616516, | |
| "learning_rate": 2.1608040201005025e-05, | |
| "loss": 1.162, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7506925207756234, | |
| "grad_norm": 0.6878906488418579, | |
| "learning_rate": 2.1105527638190957e-05, | |
| "loss": 1.0954, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7617728531855956, | |
| "grad_norm": 0.7377908825874329, | |
| "learning_rate": 2.0603015075376886e-05, | |
| "loss": 1.0891, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7728531855955678, | |
| "grad_norm": 0.6463127732276917, | |
| "learning_rate": 2.0100502512562815e-05, | |
| "loss": 1.0751, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.78393351800554, | |
| "grad_norm": 0.6408430337905884, | |
| "learning_rate": 1.9597989949748744e-05, | |
| "loss": 1.0861, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.7950138504155124, | |
| "grad_norm": 0.7149915099143982, | |
| "learning_rate": 1.9095477386934673e-05, | |
| "loss": 1.1801, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.8060941828254848, | |
| "grad_norm": 0.6683186888694763, | |
| "learning_rate": 1.8592964824120602e-05, | |
| "loss": 1.0473, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.817174515235457, | |
| "grad_norm": 0.6818321347236633, | |
| "learning_rate": 1.8090452261306535e-05, | |
| "loss": 1.0699, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.8282548476454292, | |
| "grad_norm": 0.7177061438560486, | |
| "learning_rate": 1.7587939698492464e-05, | |
| "loss": 1.12, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.8393351800554016, | |
| "grad_norm": 0.6971613168716431, | |
| "learning_rate": 1.7085427135678393e-05, | |
| "loss": 1.0919, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.850415512465374, | |
| "grad_norm": 0.6916444301605225, | |
| "learning_rate": 1.6582914572864322e-05, | |
| "loss": 1.0878, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8614958448753463, | |
| "grad_norm": 0.6586862206459045, | |
| "learning_rate": 1.608040201005025e-05, | |
| "loss": 0.9554, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8725761772853184, | |
| "grad_norm": 0.7246650457382202, | |
| "learning_rate": 1.5577889447236183e-05, | |
| "loss": 1.1607, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8836565096952909, | |
| "grad_norm": 0.761667013168335, | |
| "learning_rate": 1.507537688442211e-05, | |
| "loss": 1.1838, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.6918529868125916, | |
| "learning_rate": 1.457286432160804e-05, | |
| "loss": 1.1148, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.9058171745152355, | |
| "grad_norm": 0.7272382378578186, | |
| "learning_rate": 1.407035175879397e-05, | |
| "loss": 1.1342, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.9168975069252077, | |
| "grad_norm": 0.6923725008964539, | |
| "learning_rate": 1.3567839195979901e-05, | |
| "loss": 1.0149, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.92797783933518, | |
| "grad_norm": 0.7408672571182251, | |
| "learning_rate": 1.306532663316583e-05, | |
| "loss": 1.148, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.9390581717451525, | |
| "grad_norm": 0.6574199795722961, | |
| "learning_rate": 1.2562814070351759e-05, | |
| "loss": 1.0415, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9501385041551247, | |
| "grad_norm": 0.721842885017395, | |
| "learning_rate": 1.2060301507537688e-05, | |
| "loss": 1.2188, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9612188365650969, | |
| "grad_norm": 0.7084245681762695, | |
| "learning_rate": 1.1557788944723619e-05, | |
| "loss": 1.1464, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9722991689750693, | |
| "grad_norm": 0.6412287354469299, | |
| "learning_rate": 1.105527638190955e-05, | |
| "loss": 0.9909, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9833795013850417, | |
| "grad_norm": 0.7143056988716125, | |
| "learning_rate": 1.0552763819095479e-05, | |
| "loss": 1.1627, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.994459833795014, | |
| "grad_norm": 0.6697918772697449, | |
| "learning_rate": 1.0050251256281408e-05, | |
| "loss": 1.0892, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.005540166204986, | |
| "grad_norm": 1.8020601272583008, | |
| "learning_rate": 9.547738693467337e-06, | |
| "loss": 1.984, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.0166204986149583, | |
| "grad_norm": 0.7190349102020264, | |
| "learning_rate": 9.045226130653267e-06, | |
| "loss": 1.0837, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.027700831024931, | |
| "grad_norm": 0.6655665040016174, | |
| "learning_rate": 8.542713567839196e-06, | |
| "loss": 0.9812, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.038781163434903, | |
| "grad_norm": 0.7195249795913696, | |
| "learning_rate": 8.040201005025125e-06, | |
| "loss": 1.1748, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.0498614958448753, | |
| "grad_norm": 0.6796366572380066, | |
| "learning_rate": 7.537688442211055e-06, | |
| "loss": 1.07, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.0609418282548475, | |
| "grad_norm": 0.679729700088501, | |
| "learning_rate": 7.035175879396985e-06, | |
| "loss": 0.996, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.07202216066482, | |
| "grad_norm": 0.6805736422538757, | |
| "learning_rate": 6.532663316582915e-06, | |
| "loss": 1.0531, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0831024930747923, | |
| "grad_norm": 0.698531448841095, | |
| "learning_rate": 6.030150753768844e-06, | |
| "loss": 1.1163, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.0941828254847645, | |
| "grad_norm": 0.6734586358070374, | |
| "learning_rate": 5.527638190954775e-06, | |
| "loss": 1.0584, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.6843752264976501, | |
| "learning_rate": 5.025125628140704e-06, | |
| "loss": 1.0854, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.1163434903047094, | |
| "grad_norm": 0.6570263504981995, | |
| "learning_rate": 4.522613065326634e-06, | |
| "loss": 1.016, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.1274238227146816, | |
| "grad_norm": 0.707304060459137, | |
| "learning_rate": 4.020100502512563e-06, | |
| "loss": 1.0311, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.1385041551246537, | |
| "grad_norm": 0.7010348439216614, | |
| "learning_rate": 3.5175879396984926e-06, | |
| "loss": 1.1732, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.149584487534626, | |
| "grad_norm": 0.6604486703872681, | |
| "learning_rate": 3.015075376884422e-06, | |
| "loss": 1.012, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.160664819944598, | |
| "grad_norm": 0.6502628922462463, | |
| "learning_rate": 2.512562814070352e-06, | |
| "loss": 0.9839, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.1717451523545708, | |
| "grad_norm": 0.6600444912910461, | |
| "learning_rate": 2.0100502512562813e-06, | |
| "loss": 1.0509, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.182825484764543, | |
| "grad_norm": 0.69642573595047, | |
| "learning_rate": 1.507537688442211e-06, | |
| "loss": 1.1109, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.193905817174515, | |
| "grad_norm": 0.691315233707428, | |
| "learning_rate": 1.0050251256281407e-06, | |
| "loss": 1.1067, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.2049861495844874, | |
| "grad_norm": 0.6788798570632935, | |
| "learning_rate": 5.025125628140703e-07, | |
| "loss": 1.0564, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.21606648199446, | |
| "grad_norm": 0.6888399124145508, | |
| "learning_rate": 0.0, | |
| "loss": 1.0406, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.18753507500032e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |