| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9991537376586743, | |
| "eval_steps": 500, | |
| "global_step": 1329, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.022566995768688293, | |
| "grad_norm": 1.100711091672383, | |
| "learning_rate": 4.997501873438867e-06, | |
| "loss": 0.7439, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.045133991537376586, | |
| "grad_norm": 0.9978428347036197, | |
| "learning_rate": 4.995007487521836e-06, | |
| "loss": 0.6805, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06770098730606489, | |
| "grad_norm": 0.6349072049081254, | |
| "learning_rate": 4.992516832922945e-06, | |
| "loss": 0.663, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09026798307475317, | |
| "grad_norm": 0.6944987900561912, | |
| "learning_rate": 4.990029900348746e-06, | |
| "loss": 0.6405, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11283497884344147, | |
| "grad_norm": 0.6342717028236399, | |
| "learning_rate": 4.987546680538165e-06, | |
| "loss": 0.634, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13540197461212977, | |
| "grad_norm": 0.7029633847174581, | |
| "learning_rate": 4.985067164262359e-06, | |
| "loss": 0.6333, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15796897038081806, | |
| "grad_norm": 0.6429990155588656, | |
| "learning_rate": 4.98259134232457e-06, | |
| "loss": 0.629, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18053596614950634, | |
| "grad_norm": 0.7304393500080666, | |
| "learning_rate": 4.980119205559974e-06, | |
| "loss": 0.6305, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.20310296191819463, | |
| "grad_norm": 0.7772726554166205, | |
| "learning_rate": 4.977650744835555e-06, | |
| "loss": 0.6282, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22566995768688294, | |
| "grad_norm": 0.6015227630260552, | |
| "learning_rate": 4.975185951049947e-06, | |
| "loss": 0.6274, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24823695345557123, | |
| "grad_norm": 0.597054561284431, | |
| "learning_rate": 4.972724815133302e-06, | |
| "loss": 0.6183, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.27080394922425954, | |
| "grad_norm": 0.6470895698195512, | |
| "learning_rate": 4.970267328047151e-06, | |
| "loss": 0.6174, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2933709449929478, | |
| "grad_norm": 0.8937084791804235, | |
| "learning_rate": 4.9678134807842575e-06, | |
| "loss": 0.6186, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3159379407616361, | |
| "grad_norm": 0.6639999069888693, | |
| "learning_rate": 4.965363264368484e-06, | |
| "loss": 0.6135, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3385049365303244, | |
| "grad_norm": 0.6656810650278343, | |
| "learning_rate": 4.962916669854652e-06, | |
| "loss": 0.614, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3610719322990127, | |
| "grad_norm": 0.5853550395216752, | |
| "learning_rate": 4.960473688328407e-06, | |
| "loss": 0.616, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.383638928067701, | |
| "grad_norm": 0.6429269241253047, | |
| "learning_rate": 4.95803431090608e-06, | |
| "loss": 0.6121, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.40620592383638926, | |
| "grad_norm": 0.5934937346014221, | |
| "learning_rate": 4.955598528734554e-06, | |
| "loss": 0.6088, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4287729196050776, | |
| "grad_norm": 0.6370662067963314, | |
| "learning_rate": 4.953166332991125e-06, | |
| "loss": 0.6033, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4513399153737659, | |
| "grad_norm": 0.6456352721499912, | |
| "learning_rate": 4.950737714883372e-06, | |
| "loss": 0.6099, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.47390691114245415, | |
| "grad_norm": 0.5814512533107921, | |
| "learning_rate": 4.948312665649022e-06, | |
| "loss": 0.6063, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.49647390691114246, | |
| "grad_norm": 0.732199943548795, | |
| "learning_rate": 4.945891176555817e-06, | |
| "loss": 0.5982, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5190409026798307, | |
| "grad_norm": 0.5712191962038494, | |
| "learning_rate": 4.943473238901383e-06, | |
| "loss": 0.6069, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5416078984485191, | |
| "grad_norm": 0.6612925923310224, | |
| "learning_rate": 4.941058844013094e-06, | |
| "loss": 0.6058, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5641748942172073, | |
| "grad_norm": 0.5880543259363735, | |
| "learning_rate": 4.938647983247949e-06, | |
| "loss": 0.6, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5867418899858956, | |
| "grad_norm": 0.7220209248194411, | |
| "learning_rate": 4.936240647992436e-06, | |
| "loss": 0.6026, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.609308885754584, | |
| "grad_norm": 0.5655649436066078, | |
| "learning_rate": 4.933836829662409e-06, | |
| "loss": 0.598, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6318758815232722, | |
| "grad_norm": 0.5523686182917099, | |
| "learning_rate": 4.9314365197029475e-06, | |
| "loss": 0.6047, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6544428772919605, | |
| "grad_norm": 0.718258478379325, | |
| "learning_rate": 4.9290397095882446e-06, | |
| "loss": 0.5981, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6770098730606487, | |
| "grad_norm": 0.6923839988383336, | |
| "learning_rate": 4.9266463908214664e-06, | |
| "loss": 0.6042, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6995768688293371, | |
| "grad_norm": 0.537500360918552, | |
| "learning_rate": 4.924256554934632e-06, | |
| "loss": 0.5973, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7221438645980254, | |
| "grad_norm": 0.5213007989989484, | |
| "learning_rate": 4.9218701934884865e-06, | |
| "loss": 0.6005, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7447108603667136, | |
| "grad_norm": 0.570859069043083, | |
| "learning_rate": 4.919487298072377e-06, | |
| "loss": 0.596, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.767277856135402, | |
| "grad_norm": 0.6158715009590428, | |
| "learning_rate": 4.917107860304125e-06, | |
| "loss": 0.5955, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7898448519040903, | |
| "grad_norm": 0.6426850162244057, | |
| "learning_rate": 4.914731871829905e-06, | |
| "loss": 0.597, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8124118476727785, | |
| "grad_norm": 0.5762853195893453, | |
| "learning_rate": 4.912359324324121e-06, | |
| "loss": 0.596, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8349788434414669, | |
| "grad_norm": 0.5838318091592569, | |
| "learning_rate": 4.909990209489284e-06, | |
| "loss": 0.6031, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8575458392101551, | |
| "grad_norm": 0.5822722127512734, | |
| "learning_rate": 4.907624519055888e-06, | |
| "loss": 0.5998, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8801128349788434, | |
| "grad_norm": 0.6211815812426689, | |
| "learning_rate": 4.905262244782294e-06, | |
| "loss": 0.5938, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9026798307475318, | |
| "grad_norm": 0.5926993813948374, | |
| "learning_rate": 4.902903378454601e-06, | |
| "loss": 0.5961, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.92524682651622, | |
| "grad_norm": 0.6502046942563887, | |
| "learning_rate": 4.900547911886537e-06, | |
| "loss": 0.594, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9478138222849083, | |
| "grad_norm": 0.6150531879942983, | |
| "learning_rate": 4.898195836919327e-06, | |
| "loss": 0.5947, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9703808180535967, | |
| "grad_norm": 0.5693169684981371, | |
| "learning_rate": 4.895847145421587e-06, | |
| "loss": 0.5889, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9929478138222849, | |
| "grad_norm": 0.6633453342653501, | |
| "learning_rate": 4.893501829289195e-06, | |
| "loss": 0.5937, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9997179125528914, | |
| "eval_loss": 0.5915236473083496, | |
| "eval_runtime": 687.4788, | |
| "eval_samples_per_second": 17.369, | |
| "eval_steps_per_second": 0.544, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.0155148095909732, | |
| "grad_norm": 0.7247787853500239, | |
| "learning_rate": 4.891159880445185e-06, | |
| "loss": 0.6015, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0380818053596614, | |
| "grad_norm": 0.6068919916237175, | |
| "learning_rate": 4.888821290839617e-06, | |
| "loss": 0.5289, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0606488011283497, | |
| "grad_norm": 0.60326825190498, | |
| "learning_rate": 4.886486052449469e-06, | |
| "loss": 0.5437, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0832157968970382, | |
| "grad_norm": 0.5689189414243914, | |
| "learning_rate": 4.8841541572785224e-06, | |
| "loss": 0.5316, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1057827926657264, | |
| "grad_norm": 0.6385824037992914, | |
| "learning_rate": 4.881825597357242e-06, | |
| "loss": 0.5417, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.1283497884344147, | |
| "grad_norm": 0.5854327072193312, | |
| "learning_rate": 4.8795003647426654e-06, | |
| "loss": 0.5404, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.150916784203103, | |
| "grad_norm": 0.5947076927155363, | |
| "learning_rate": 4.877178451518289e-06, | |
| "loss": 0.5415, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.1734837799717912, | |
| "grad_norm": 0.5719055022308829, | |
| "learning_rate": 4.8748598497939494e-06, | |
| "loss": 0.534, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1960507757404795, | |
| "grad_norm": 0.6903559934708517, | |
| "learning_rate": 4.872544551705718e-06, | |
| "loss": 0.5404, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.2186177715091677, | |
| "grad_norm": 0.5635637424337575, | |
| "learning_rate": 4.870232549415787e-06, | |
| "loss": 0.5379, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.2411847672778562, | |
| "grad_norm": 0.596695858753005, | |
| "learning_rate": 4.867923835112355e-06, | |
| "loss": 0.5427, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.2637517630465445, | |
| "grad_norm": 0.6228671033057678, | |
| "learning_rate": 4.865618401009519e-06, | |
| "loss": 0.5443, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2863187588152327, | |
| "grad_norm": 0.5663816262212089, | |
| "learning_rate": 4.863316239347163e-06, | |
| "loss": 0.5416, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.308885754583921, | |
| "grad_norm": 0.6145942738735419, | |
| "learning_rate": 4.861017342390847e-06, | |
| "loss": 0.5417, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.3314527503526092, | |
| "grad_norm": 0.5647488472588417, | |
| "learning_rate": 4.858721702431704e-06, | |
| "loss": 0.5417, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.3540197461212977, | |
| "grad_norm": 0.5686060037431164, | |
| "learning_rate": 4.856429311786322e-06, | |
| "loss": 0.5469, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.376586741889986, | |
| "grad_norm": 0.5476278402050497, | |
| "learning_rate": 4.8541401627966426e-06, | |
| "loss": 0.5395, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3991537376586742, | |
| "grad_norm": 0.5651482323611746, | |
| "learning_rate": 4.85185424782985e-06, | |
| "loss": 0.5437, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.4217207334273625, | |
| "grad_norm": 0.5818373440698466, | |
| "learning_rate": 4.8495715592782715e-06, | |
| "loss": 0.5417, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.4442877291960508, | |
| "grad_norm": 0.6943754325038471, | |
| "learning_rate": 4.847292089559258e-06, | |
| "loss": 0.5415, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.466854724964739, | |
| "grad_norm": 0.5809337717523729, | |
| "learning_rate": 4.845015831115093e-06, | |
| "loss": 0.5422, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.4894217207334273, | |
| "grad_norm": 0.6366781600380568, | |
| "learning_rate": 4.842742776412874e-06, | |
| "loss": 0.546, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.5119887165021155, | |
| "grad_norm": 0.5848333620078674, | |
| "learning_rate": 4.840472917944417e-06, | |
| "loss": 0.5407, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.5345557122708038, | |
| "grad_norm": 0.5539620278393851, | |
| "learning_rate": 4.838206248226147e-06, | |
| "loss": 0.5348, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.5571227080394923, | |
| "grad_norm": 0.595515096482354, | |
| "learning_rate": 4.835942759799002e-06, | |
| "loss": 0.537, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.5796897038081805, | |
| "grad_norm": 0.5765265849661396, | |
| "learning_rate": 4.833682445228318e-06, | |
| "loss": 0.5411, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.6022566995768688, | |
| "grad_norm": 0.5571197050761847, | |
| "learning_rate": 4.831425297103738e-06, | |
| "loss": 0.5438, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.6248236953455573, | |
| "grad_norm": 0.5792740919003788, | |
| "learning_rate": 4.829171308039099e-06, | |
| "loss": 0.5518, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.6473906911142455, | |
| "grad_norm": 0.5636302795085814, | |
| "learning_rate": 4.826920470672344e-06, | |
| "loss": 0.5352, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.6699576868829338, | |
| "grad_norm": 0.7262762150102657, | |
| "learning_rate": 4.824672777665406e-06, | |
| "loss": 0.5511, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.692524682651622, | |
| "grad_norm": 0.6203455822074458, | |
| "learning_rate": 4.822428221704122e-06, | |
| "loss": 0.5423, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.7150916784203103, | |
| "grad_norm": 0.6314021891275363, | |
| "learning_rate": 4.820186795498119e-06, | |
| "loss": 0.5413, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.7376586741889986, | |
| "grad_norm": 0.6244754158170291, | |
| "learning_rate": 4.817948491780728e-06, | |
| "loss": 0.5419, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.7602256699576868, | |
| "grad_norm": 0.6285912361574367, | |
| "learning_rate": 4.815713303308872e-06, | |
| "loss": 0.5377, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.782792665726375, | |
| "grad_norm": 0.5663852190944293, | |
| "learning_rate": 4.813481222862981e-06, | |
| "loss": 0.5414, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.8053596614950633, | |
| "grad_norm": 0.6930428334287735, | |
| "learning_rate": 4.811252243246881e-06, | |
| "loss": 0.5365, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.8279266572637518, | |
| "grad_norm": 0.524109251310443, | |
| "learning_rate": 4.809026357287709e-06, | |
| "loss": 0.547, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.85049365303244, | |
| "grad_norm": 0.6577534946547546, | |
| "learning_rate": 4.806803557835802e-06, | |
| "loss": 0.5343, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.8730606488011283, | |
| "grad_norm": 0.561047334945624, | |
| "learning_rate": 4.804583837764616e-06, | |
| "loss": 0.5463, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.8956276445698168, | |
| "grad_norm": 0.5709254970367929, | |
| "learning_rate": 4.802367189970616e-06, | |
| "loss": 0.5463, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.918194640338505, | |
| "grad_norm": 0.5539876831180034, | |
| "learning_rate": 4.8001536073731936e-06, | |
| "loss": 0.5342, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.9407616361071933, | |
| "grad_norm": 0.6261176438608845, | |
| "learning_rate": 4.797943082914558e-06, | |
| "loss": 0.5393, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.9633286318758816, | |
| "grad_norm": 0.6784873891334788, | |
| "learning_rate": 4.795735609559657e-06, | |
| "loss": 0.5457, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.9858956276445698, | |
| "grad_norm": 0.6032255886331123, | |
| "learning_rate": 4.793531180296065e-06, | |
| "loss": 0.5439, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.9994358251057829, | |
| "eval_loss": 0.5869531631469727, | |
| "eval_runtime": 690.9414, | |
| "eval_samples_per_second": 17.282, | |
| "eval_steps_per_second": 0.541, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 2.008462623413258, | |
| "grad_norm": 0.9949702529495333, | |
| "learning_rate": 4.7913297881339085e-06, | |
| "loss": 0.5679, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.0310296191819464, | |
| "grad_norm": 0.7201678126466289, | |
| "learning_rate": 4.789131426105757e-06, | |
| "loss": 0.4798, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.0535966149506346, | |
| "grad_norm": 0.674929621297509, | |
| "learning_rate": 4.786936087266542e-06, | |
| "loss": 0.4817, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.076163610719323, | |
| "grad_norm": 0.6070882301942089, | |
| "learning_rate": 4.784743764693455e-06, | |
| "loss": 0.4741, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.098730606488011, | |
| "grad_norm": 0.6063712760118354, | |
| "learning_rate": 4.7825544514858655e-06, | |
| "loss": 0.4814, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.1212976022566994, | |
| "grad_norm": 0.5947320209763772, | |
| "learning_rate": 4.780368140765222e-06, | |
| "loss": 0.4856, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.143864598025388, | |
| "grad_norm": 0.6280872907961158, | |
| "learning_rate": 4.778184825674966e-06, | |
| "loss": 0.4873, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.1664315937940763, | |
| "grad_norm": 0.6188168725190593, | |
| "learning_rate": 4.776004499380439e-06, | |
| "loss": 0.4837, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.1889985895627646, | |
| "grad_norm": 0.7603271106624941, | |
| "learning_rate": 4.773827155068793e-06, | |
| "loss": 0.4818, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.211565585331453, | |
| "grad_norm": 0.5975717089201273, | |
| "learning_rate": 4.771652785948902e-06, | |
| "loss": 0.4767, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.234132581100141, | |
| "grad_norm": 0.5809905836332211, | |
| "learning_rate": 4.769481385251275e-06, | |
| "loss": 0.4765, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.2566995768688294, | |
| "grad_norm": 0.5843693488329456, | |
| "learning_rate": 4.767312946227961e-06, | |
| "loss": 0.4855, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.2792665726375176, | |
| "grad_norm": 0.6027690911140321, | |
| "learning_rate": 4.765147462152471e-06, | |
| "loss": 0.4919, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.301833568406206, | |
| "grad_norm": 0.7192183771609572, | |
| "learning_rate": 4.762984926319677e-06, | |
| "loss": 0.4861, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.324400564174894, | |
| "grad_norm": 0.645690672789835, | |
| "learning_rate": 4.760825332045738e-06, | |
| "loss": 0.4905, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.3469675599435824, | |
| "grad_norm": 0.6311156323408128, | |
| "learning_rate": 4.758668672668006e-06, | |
| "loss": 0.4883, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.3695345557122707, | |
| "grad_norm": 0.6045990122500915, | |
| "learning_rate": 4.756514941544941e-06, | |
| "loss": 0.4861, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.392101551480959, | |
| "grad_norm": 0.5914240795148143, | |
| "learning_rate": 4.754364132056025e-06, | |
| "loss": 0.4855, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.414668547249647, | |
| "grad_norm": 0.6967797148982833, | |
| "learning_rate": 4.752216237601676e-06, | |
| "loss": 0.4869, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.4372355430183354, | |
| "grad_norm": 0.7069219663540367, | |
| "learning_rate": 4.750071251603165e-06, | |
| "loss": 0.49, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.459802538787024, | |
| "grad_norm": 0.593026634099332, | |
| "learning_rate": 4.7479291675025314e-06, | |
| "loss": 0.488, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.4823695345557124, | |
| "grad_norm": 0.6133731329239802, | |
| "learning_rate": 4.745789978762496e-06, | |
| "loss": 0.491, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.5049365303244007, | |
| "grad_norm": 0.5823330923339051, | |
| "learning_rate": 4.7436536788663765e-06, | |
| "loss": 0.4943, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.527503526093089, | |
| "grad_norm": 0.6537812247886863, | |
| "learning_rate": 4.74152026131801e-06, | |
| "loss": 0.4848, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.550070521861777, | |
| "grad_norm": 0.6021010883789748, | |
| "learning_rate": 4.739389719641665e-06, | |
| "loss": 0.4899, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.5726375176304654, | |
| "grad_norm": 0.5782734444486691, | |
| "learning_rate": 4.7372620473819615e-06, | |
| "loss": 0.4883, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.5952045133991537, | |
| "grad_norm": 0.6403059046505779, | |
| "learning_rate": 4.735137238103785e-06, | |
| "loss": 0.489, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.617771509167842, | |
| "grad_norm": 0.6255079366673328, | |
| "learning_rate": 4.7330152853922064e-06, | |
| "loss": 0.4927, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.64033850493653, | |
| "grad_norm": 0.6422810171594653, | |
| "learning_rate": 4.730896182852409e-06, | |
| "loss": 0.4817, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.6629055007052185, | |
| "grad_norm": 0.6348937375312369, | |
| "learning_rate": 4.72877992410959e-06, | |
| "loss": 0.4924, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.685472496473907, | |
| "grad_norm": 0.6020419858221361, | |
| "learning_rate": 4.7266665028088985e-06, | |
| "loss": 0.4932, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.7080394922425954, | |
| "grad_norm": 0.5849628799150851, | |
| "learning_rate": 4.72455591261534e-06, | |
| "loss": 0.4917, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.7306064880112837, | |
| "grad_norm": 0.5754324505606723, | |
| "learning_rate": 4.722448147213712e-06, | |
| "loss": 0.493, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.753173483779972, | |
| "grad_norm": 0.5734988589501125, | |
| "learning_rate": 4.720343200308507e-06, | |
| "loss": 0.4896, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.77574047954866, | |
| "grad_norm": 0.5710201726782703, | |
| "learning_rate": 4.7182410656238484e-06, | |
| "loss": 0.4961, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.7983074753173485, | |
| "grad_norm": 0.609980767169798, | |
| "learning_rate": 4.716141736903407e-06, | |
| "loss": 0.4944, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.8208744710860367, | |
| "grad_norm": 0.5680473087970347, | |
| "learning_rate": 4.714045207910318e-06, | |
| "loss": 0.4945, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.843441466854725, | |
| "grad_norm": 0.5545814696527704, | |
| "learning_rate": 4.71195147242711e-06, | |
| "loss": 0.4963, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.8660084626234132, | |
| "grad_norm": 0.5701759707556856, | |
| "learning_rate": 4.709860524255622e-06, | |
| "loss": 0.4932, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.8885754583921015, | |
| "grad_norm": 0.5598897964453792, | |
| "learning_rate": 4.707772357216934e-06, | |
| "loss": 0.4912, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.9111424541607898, | |
| "grad_norm": 0.6106914180921081, | |
| "learning_rate": 4.705686965151282e-06, | |
| "loss": 0.4949, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.933709449929478, | |
| "grad_norm": 0.6259987776719272, | |
| "learning_rate": 4.703604341917987e-06, | |
| "loss": 0.4908, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.9562764456981663, | |
| "grad_norm": 0.6635742267932128, | |
| "learning_rate": 4.701524481395374e-06, | |
| "loss": 0.4946, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.9788434414668545, | |
| "grad_norm": 0.5993621544317691, | |
| "learning_rate": 4.699447377480703e-06, | |
| "loss": 0.4909, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.9991537376586743, | |
| "eval_loss": 0.5994867086410522, | |
| "eval_runtime": 687.8101, | |
| "eval_samples_per_second": 17.361, | |
| "eval_steps_per_second": 0.544, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 2.9991537376586743, | |
| "step": 1329, | |
| "total_flos": 5064195066298368.0, | |
| "train_loss": 0.5484652002562788, | |
| "train_runtime": 120694.0629, | |
| "train_samples_per_second": 5.639, | |
| "train_steps_per_second": 0.011 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1329, | |
| "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": 5064195066298368.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |