| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9878213802435725, | |
| "eval_steps": 500, | |
| "global_step": 276, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010825439783491205, | |
| "grad_norm": 5.808782195509016, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.8566, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02165087956698241, | |
| "grad_norm": 5.813205553053294, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.8603, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03247631935047361, | |
| "grad_norm": 5.437266634937425, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.8524, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04330175913396482, | |
| "grad_norm": 2.304459428183892, | |
| "learning_rate": 1.1428571428571429e-05, | |
| "loss": 0.7759, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05412719891745602, | |
| "grad_norm": 3.927953214676238, | |
| "learning_rate": 1.4285714285714287e-05, | |
| "loss": 0.7544, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06495263870094722, | |
| "grad_norm": 4.210566313942864, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 0.751, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07577807848443843, | |
| "grad_norm": 4.355428233675389, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7344, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08660351826792964, | |
| "grad_norm": 2.9135046061222485, | |
| "learning_rate": 2.2857142857142858e-05, | |
| "loss": 0.7289, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09742895805142084, | |
| "grad_norm": 2.7435977316235043, | |
| "learning_rate": 2.5714285714285718e-05, | |
| "loss": 0.672, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10825439783491204, | |
| "grad_norm": 2.124176021674874, | |
| "learning_rate": 2.8571428571428574e-05, | |
| "loss": 0.6658, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11907983761840325, | |
| "grad_norm": 1.560974058082246, | |
| "learning_rate": 3.142857142857143e-05, | |
| "loss": 0.6378, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12990527740189445, | |
| "grad_norm": 1.356421259435607, | |
| "learning_rate": 3.4285714285714284e-05, | |
| "loss": 0.6303, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14073071718538566, | |
| "grad_norm": 1.5660206311423615, | |
| "learning_rate": 3.714285714285715e-05, | |
| "loss": 0.614, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15155615696887687, | |
| "grad_norm": 1.2014017650689, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6243, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16238159675236807, | |
| "grad_norm": 1.4027226467953162, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 0.6077, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17320703653585928, | |
| "grad_norm": 1.1683655505452553, | |
| "learning_rate": 4.5714285714285716e-05, | |
| "loss": 0.5975, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.18403247631935046, | |
| "grad_norm": 1.835593785225974, | |
| "learning_rate": 4.857142857142857e-05, | |
| "loss": 0.596, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.19485791610284167, | |
| "grad_norm": 1.1345154883647737, | |
| "learning_rate": 5.1428571428571436e-05, | |
| "loss": 0.5762, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20568335588633288, | |
| "grad_norm": 1.824887208898125, | |
| "learning_rate": 5.4285714285714295e-05, | |
| "loss": 0.572, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2165087956698241, | |
| "grad_norm": 1.6126198323929144, | |
| "learning_rate": 5.714285714285715e-05, | |
| "loss": 0.5762, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2273342354533153, | |
| "grad_norm": 1.5918620480095331, | |
| "learning_rate": 6.000000000000001e-05, | |
| "loss": 0.5692, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2381596752368065, | |
| "grad_norm": 1.4250264018215846, | |
| "learning_rate": 6.285714285714286e-05, | |
| "loss": 0.5693, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2489851150202977, | |
| "grad_norm": 1.6782182483587103, | |
| "learning_rate": 6.571428571428571e-05, | |
| "loss": 0.5601, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2598105548037889, | |
| "grad_norm": 1.2505969451610879, | |
| "learning_rate": 6.857142857142857e-05, | |
| "loss": 0.5655, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2706359945872801, | |
| "grad_norm": 2.1118955155765673, | |
| "learning_rate": 7.142857142857143e-05, | |
| "loss": 0.5609, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2814614343707713, | |
| "grad_norm": 1.4121413273832522, | |
| "learning_rate": 7.42857142857143e-05, | |
| "loss": 0.5389, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2922868741542625, | |
| "grad_norm": 2.0766991334479408, | |
| "learning_rate": 7.714285714285715e-05, | |
| "loss": 0.547, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.30311231393775373, | |
| "grad_norm": 1.4556064554736243, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5524, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.31393775372124494, | |
| "grad_norm": 1.6077536748554087, | |
| "learning_rate": 7.999679062421315e-05, | |
| "loss": 0.5492, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.32476319350473615, | |
| "grad_norm": 1.5101627731283567, | |
| "learning_rate": 7.998716301185722e-05, | |
| "loss": 0.5436, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33558863328822736, | |
| "grad_norm": 1.984676597011515, | |
| "learning_rate": 7.997111870786354e-05, | |
| "loss": 0.5451, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34641407307171856, | |
| "grad_norm": 1.21014550417124, | |
| "learning_rate": 7.994866028684212e-05, | |
| "loss": 0.5388, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3572395128552097, | |
| "grad_norm": 1.95150099944423, | |
| "learning_rate": 7.991979135266861e-05, | |
| "loss": 0.5414, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3680649526387009, | |
| "grad_norm": 1.5787618301801385, | |
| "learning_rate": 7.98845165379059e-05, | |
| "loss": 0.5454, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.37889039242219213, | |
| "grad_norm": 1.0690263609490216, | |
| "learning_rate": 7.984284150306085e-05, | |
| "loss": 0.5239, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.38971583220568334, | |
| "grad_norm": 1.7977915032515326, | |
| "learning_rate": 7.97947729356758e-05, | |
| "loss": 0.5253, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.40054127198917455, | |
| "grad_norm": 2.120705589307475, | |
| "learning_rate": 7.974031854925562e-05, | |
| "loss": 0.535, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41136671177266576, | |
| "grad_norm": 1.1488904564251323, | |
| "learning_rate": 7.967948708202972e-05, | |
| "loss": 0.5189, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.42219215155615697, | |
| "grad_norm": 2.7953953760475363, | |
| "learning_rate": 7.961228829555003e-05, | |
| "loss": 0.5404, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.4330175913396482, | |
| "grad_norm": 2.211003461793142, | |
| "learning_rate": 7.953873297312447e-05, | |
| "loss": 0.5388, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4438430311231394, | |
| "grad_norm": 1.6204765603156832, | |
| "learning_rate": 7.945883291808655e-05, | |
| "loss": 0.5234, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4546684709066306, | |
| "grad_norm": 1.560460098341002, | |
| "learning_rate": 7.937260095190137e-05, | |
| "loss": 0.5191, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4654939106901218, | |
| "grad_norm": 1.3057494076782257, | |
| "learning_rate": 7.928005091210817e-05, | |
| "loss": 0.514, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.476319350473613, | |
| "grad_norm": 1.3681049413609492, | |
| "learning_rate": 7.918119765009979e-05, | |
| "loss": 0.5172, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4871447902571042, | |
| "grad_norm": 1.3087285372962363, | |
| "learning_rate": 7.907605702873948e-05, | |
| "loss": 0.5166, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4979702300405954, | |
| "grad_norm": 1.5892591530023001, | |
| "learning_rate": 7.896464591981549e-05, | |
| "loss": 0.5135, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5087956698240866, | |
| "grad_norm": 0.9146670461290698, | |
| "learning_rate": 7.884698220133357e-05, | |
| "loss": 0.5093, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5196211096075778, | |
| "grad_norm": 1.2871074236378093, | |
| "learning_rate": 7.872308475464818e-05, | |
| "loss": 0.5066, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.530446549391069, | |
| "grad_norm": 1.187151202508197, | |
| "learning_rate": 7.859297346143258e-05, | |
| "loss": 0.5058, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5412719891745602, | |
| "grad_norm": 1.5765769924422004, | |
| "learning_rate": 7.84566692004885e-05, | |
| "loss": 0.5027, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5520974289580515, | |
| "grad_norm": 1.1216495666926116, | |
| "learning_rate": 7.831419384439565e-05, | |
| "loss": 0.4992, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5629228687415426, | |
| "grad_norm": 1.22301226354295, | |
| "learning_rate": 7.816557025600196e-05, | |
| "loss": 0.494, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5737483085250338, | |
| "grad_norm": 1.164455112281458, | |
| "learning_rate": 7.80108222847547e-05, | |
| "loss": 0.4938, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.584573748308525, | |
| "grad_norm": 1.4650064133181142, | |
| "learning_rate": 7.784997476287349e-05, | |
| "loss": 0.5074, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5953991880920162, | |
| "grad_norm": 1.22737056565541, | |
| "learning_rate": 7.76830535013654e-05, | |
| "loss": 0.494, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6062246278755075, | |
| "grad_norm": 0.8673329147060569, | |
| "learning_rate": 7.751008528588322e-05, | |
| "loss": 0.4958, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6170500676589986, | |
| "grad_norm": 1.427111220332804, | |
| "learning_rate": 7.733109787242708e-05, | |
| "loss": 0.5, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6278755074424899, | |
| "grad_norm": 1.2086609324924107, | |
| "learning_rate": 7.71461199828905e-05, | |
| "loss": 0.4938, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.638700947225981, | |
| "grad_norm": 1.2921503295566135, | |
| "learning_rate": 7.695518130045147e-05, | |
| "loss": 0.4841, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6495263870094723, | |
| "grad_norm": 1.2045601060877222, | |
| "learning_rate": 7.675831246480923e-05, | |
| "loss": 0.4928, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6603518267929634, | |
| "grad_norm": 1.2673683923210595, | |
| "learning_rate": 7.655554506726747e-05, | |
| "loss": 0.502, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6711772665764547, | |
| "grad_norm": 1.1256001884788949, | |
| "learning_rate": 7.6346911645665e-05, | |
| "loss": 0.4889, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6820027063599459, | |
| "grad_norm": 0.8830494632014936, | |
| "learning_rate": 7.61324456791544e-05, | |
| "loss": 0.4955, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6928281461434371, | |
| "grad_norm": 0.9599498090811002, | |
| "learning_rate": 7.591218158282968e-05, | |
| "loss": 0.491, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7036535859269283, | |
| "grad_norm": 1.1534591144602964, | |
| "learning_rate": 7.568615470220369e-05, | |
| "loss": 0.4882, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7144790257104194, | |
| "grad_norm": 1.5164318002685795, | |
| "learning_rate": 7.545440130753634e-05, | |
| "loss": 0.4939, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7253044654939107, | |
| "grad_norm": 0.6951797833665687, | |
| "learning_rate": 7.52169585880143e-05, | |
| "loss": 0.4885, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7361299052774019, | |
| "grad_norm": 1.2588621776669873, | |
| "learning_rate": 7.497386464578329e-05, | |
| "loss": 0.5013, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7469553450608931, | |
| "grad_norm": 1.4705912459647563, | |
| "learning_rate": 7.472515848983394e-05, | |
| "loss": 0.4907, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7577807848443843, | |
| "grad_norm": 0.6781089249417709, | |
| "learning_rate": 7.447088002974199e-05, | |
| "loss": 0.4737, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7686062246278755, | |
| "grad_norm": 1.3533306719892424, | |
| "learning_rate": 7.421107006926408e-05, | |
| "loss": 0.4795, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7794316644113667, | |
| "grad_norm": 1.0711976523099835, | |
| "learning_rate": 7.394577029979004e-05, | |
| "loss": 0.4878, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.790257104194858, | |
| "grad_norm": 1.7828745135486515, | |
| "learning_rate": 7.367502329365268e-05, | |
| "loss": 0.4975, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8010825439783491, | |
| "grad_norm": 0.9406459074769725, | |
| "learning_rate": 7.33988724972963e-05, | |
| "loss": 0.4912, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8119079837618404, | |
| "grad_norm": 1.350705792808143, | |
| "learning_rate": 7.311736222430487e-05, | |
| "loss": 0.4909, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8227334235453315, | |
| "grad_norm": 0.7667816696092303, | |
| "learning_rate": 7.283053764829106e-05, | |
| "loss": 0.48, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8335588633288228, | |
| "grad_norm": 1.3313352979799213, | |
| "learning_rate": 7.253844479564737e-05, | |
| "loss": 0.4817, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8443843031123139, | |
| "grad_norm": 0.7190091824384284, | |
| "learning_rate": 7.224113053816021e-05, | |
| "loss": 0.4839, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8552097428958051, | |
| "grad_norm": 1.09462918815019, | |
| "learning_rate": 7.193864258548855e-05, | |
| "loss": 0.489, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8660351826792964, | |
| "grad_norm": 1.0916277655373101, | |
| "learning_rate": 7.163102947750794e-05, | |
| "loss": 0.483, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8768606224627875, | |
| "grad_norm": 1.390674863582184, | |
| "learning_rate": 7.131834057652142e-05, | |
| "loss": 0.4911, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8876860622462788, | |
| "grad_norm": 0.8966944634191916, | |
| "learning_rate": 7.100062605933835e-05, | |
| "loss": 0.48, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8985115020297699, | |
| "grad_norm": 0.8382478308469653, | |
| "learning_rate": 7.067793690922268e-05, | |
| "loss": 0.4811, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9093369418132612, | |
| "grad_norm": 0.8056638959260214, | |
| "learning_rate": 7.035032490771165e-05, | |
| "loss": 0.483, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9201623815967523, | |
| "grad_norm": 0.8261415511137054, | |
| "learning_rate": 7.001784262630652e-05, | |
| "loss": 0.478, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9309878213802436, | |
| "grad_norm": 0.8582439051300673, | |
| "learning_rate": 6.968054341803644e-05, | |
| "loss": 0.4837, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9418132611637348, | |
| "grad_norm": 1.1350401399820995, | |
| "learning_rate": 6.933848140889705e-05, | |
| "loss": 0.4803, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.952638700947226, | |
| "grad_norm": 1.48619526541436, | |
| "learning_rate": 6.89917114891648e-05, | |
| "loss": 0.4687, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.9634641407307172, | |
| "grad_norm": 0.742051335326601, | |
| "learning_rate": 6.864028930458892e-05, | |
| "loss": 0.4788, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.9742895805142084, | |
| "grad_norm": 0.6782228718743897, | |
| "learning_rate": 6.828427124746191e-05, | |
| "loss": 0.4775, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9851150202976996, | |
| "grad_norm": 1.462261986108606, | |
| "learning_rate": 6.792371444757037e-05, | |
| "loss": 0.4848, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.9959404600811907, | |
| "grad_norm": 1.1720362746872328, | |
| "learning_rate": 6.755867676302747e-05, | |
| "loss": 0.4811, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.006765899864682, | |
| "grad_norm": 1.3982192683722436, | |
| "learning_rate": 6.718921677098853e-05, | |
| "loss": 0.7468, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0175913396481733, | |
| "grad_norm": 1.6335922067721735, | |
| "learning_rate": 6.681539375825115e-05, | |
| "loss": 0.4665, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0284167794316643, | |
| "grad_norm": 0.8967439830099186, | |
| "learning_rate": 6.643726771174164e-05, | |
| "loss": 0.4556, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0392422192151556, | |
| "grad_norm": 1.7895944548607083, | |
| "learning_rate": 6.60548993088889e-05, | |
| "loss": 0.4659, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0500676589986468, | |
| "grad_norm": 1.168440641440536, | |
| "learning_rate": 6.56683499078876e-05, | |
| "loss": 0.4487, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.060893098782138, | |
| "grad_norm": 1.7145903444920905, | |
| "learning_rate": 6.527768153785216e-05, | |
| "loss": 0.4679, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0717185385656292, | |
| "grad_norm": 1.7706267980698807, | |
| "learning_rate": 6.488295688886295e-05, | |
| "loss": 0.4602, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.0825439783491204, | |
| "grad_norm": 0.8279569864134981, | |
| "learning_rate": 6.448423930190653e-05, | |
| "loss": 0.451, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0933694181326117, | |
| "grad_norm": 1.4178241026761327, | |
| "learning_rate": 6.408159275871132e-05, | |
| "loss": 0.4524, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.104194857916103, | |
| "grad_norm": 0.855662456588115, | |
| "learning_rate": 6.36750818714807e-05, | |
| "loss": 0.4501, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.115020297699594, | |
| "grad_norm": 1.2774070048056905, | |
| "learning_rate": 6.326477187252455e-05, | |
| "loss": 0.4562, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1258457374830853, | |
| "grad_norm": 0.9647061679734499, | |
| "learning_rate": 6.28507286037917e-05, | |
| "loss": 0.4455, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1366711772665765, | |
| "grad_norm": 0.8152001814683185, | |
| "learning_rate": 6.243301850630419e-05, | |
| "loss": 0.4498, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1474966170500678, | |
| "grad_norm": 0.8148817931032433, | |
| "learning_rate": 6.201170860949565e-05, | |
| "loss": 0.4526, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1583220568335588, | |
| "grad_norm": 0.7054191999169396, | |
| "learning_rate": 6.15868665204552e-05, | |
| "loss": 0.4469, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.16914749661705, | |
| "grad_norm": 0.7721709008206348, | |
| "learning_rate": 6.11585604130785e-05, | |
| "loss": 0.4414, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1799729364005414, | |
| "grad_norm": 0.7332249667079027, | |
| "learning_rate": 6.072685901712808e-05, | |
| "loss": 0.4468, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1907983761840324, | |
| "grad_norm": 0.7820399298347422, | |
| "learning_rate": 6.02918316072043e-05, | |
| "loss": 0.4519, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2016238159675237, | |
| "grad_norm": 0.6492150545629941, | |
| "learning_rate": 5.9853547991628967e-05, | |
| "loss": 0.4498, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.212449255751015, | |
| "grad_norm": 0.6404997166391087, | |
| "learning_rate": 5.941207850124325e-05, | |
| "loss": 0.4433, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2232746955345062, | |
| "grad_norm": 0.573153411012465, | |
| "learning_rate": 5.896749397812181e-05, | |
| "loss": 0.4393, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2341001353179972, | |
| "grad_norm": 0.48292527952470066, | |
| "learning_rate": 5.8519865764204834e-05, | |
| "loss": 0.4463, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2449255751014885, | |
| "grad_norm": 0.5776286655791839, | |
| "learning_rate": 5.8069265689849884e-05, | |
| "loss": 0.4519, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2557510148849798, | |
| "grad_norm": 0.4552967048568162, | |
| "learning_rate": 5.761576606230538e-05, | |
| "loss": 0.4438, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.266576454668471, | |
| "grad_norm": 0.4974927705723956, | |
| "learning_rate": 5.7159439654107506e-05, | |
| "loss": 0.4409, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.277401894451962, | |
| "grad_norm": 0.4434973022939995, | |
| "learning_rate": 5.6700359691402533e-05, | |
| "loss": 0.441, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2882273342354533, | |
| "grad_norm": 0.45448947368164877, | |
| "learning_rate": 5.6238599842196285e-05, | |
| "loss": 0.4449, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.2990527740189446, | |
| "grad_norm": 0.4481602723894844, | |
| "learning_rate": 5.5774234204532746e-05, | |
| "loss": 0.4368, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3098782138024356, | |
| "grad_norm": 0.4642402176711816, | |
| "learning_rate": 5.5307337294603595e-05, | |
| "loss": 0.4439, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.320703653585927, | |
| "grad_norm": 0.38843059591219814, | |
| "learning_rate": 5.483798403479072e-05, | |
| "loss": 0.4395, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3315290933694182, | |
| "grad_norm": 0.37537289111217187, | |
| "learning_rate": 5.436624974164349e-05, | |
| "loss": 0.4426, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3423545331529092, | |
| "grad_norm": 0.33092526529989286, | |
| "learning_rate": 5.389221011379281e-05, | |
| "loss": 0.4397, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3531799729364005, | |
| "grad_norm": 0.3372237449147885, | |
| "learning_rate": 5.3415941219803895e-05, | |
| "loss": 0.44, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3640054127198917, | |
| "grad_norm": 0.33360582606351236, | |
| "learning_rate": 5.2937519485969525e-05, | |
| "loss": 0.4392, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.374830852503383, | |
| "grad_norm": 0.32080810729314085, | |
| "learning_rate": 5.245702168404616e-05, | |
| "loss": 0.4321, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3856562922868743, | |
| "grad_norm": 0.33377674127634516, | |
| "learning_rate": 5.1974524918934336e-05, | |
| "loss": 0.4347, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3964817320703653, | |
| "grad_norm": 0.3265043304083045, | |
| "learning_rate": 5.14901066163058e-05, | |
| "loss": 0.4338, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4073071718538566, | |
| "grad_norm": 0.405023433971877, | |
| "learning_rate": 5.1003844510179126e-05, | |
| "loss": 0.4385, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4181326116373478, | |
| "grad_norm": 0.358827748620467, | |
| "learning_rate": 5.0515816630445795e-05, | |
| "loss": 0.4403, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4289580514208389, | |
| "grad_norm": 0.3971700589793911, | |
| "learning_rate": 5.002610129034883e-05, | |
| "loss": 0.4416, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4397834912043301, | |
| "grad_norm": 0.5549803120974646, | |
| "learning_rate": 4.953477707391597e-05, | |
| "loss": 0.4353, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4506089309878214, | |
| "grad_norm": 0.7247883056299059, | |
| "learning_rate": 4.90419228233494e-05, | |
| "loss": 0.4374, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4614343707713127, | |
| "grad_norm": 0.7920917923543059, | |
| "learning_rate": 4.854761762637403e-05, | |
| "loss": 0.4379, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.472259810554804, | |
| "grad_norm": 0.8481877343768299, | |
| "learning_rate": 4.805194080354641e-05, | |
| "loss": 0.4371, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.483085250338295, | |
| "grad_norm": 0.8347261592537676, | |
| "learning_rate": 4.7554971895526175e-05, | |
| "loss": 0.4338, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4939106901217862, | |
| "grad_norm": 0.7621817880915904, | |
| "learning_rate": 4.705679065031235e-05, | |
| "loss": 0.4402, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5047361299052775, | |
| "grad_norm": 0.5219445352980245, | |
| "learning_rate": 4.6557477010446206e-05, | |
| "loss": 0.4389, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5155615696887685, | |
| "grad_norm": 0.24046891519988917, | |
| "learning_rate": 4.605711110018307e-05, | |
| "loss": 0.4379, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5263870094722598, | |
| "grad_norm": 0.5809161771814877, | |
| "learning_rate": 4.555577321263477e-05, | |
| "loss": 0.4345, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.537212449255751, | |
| "grad_norm": 0.7872874905802484, | |
| "learning_rate": 4.505354379688518e-05, | |
| "loss": 0.4355, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.548037889039242, | |
| "grad_norm": 0.6611587307034023, | |
| "learning_rate": 4.4550503445080606e-05, | |
| "loss": 0.4416, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5588633288227334, | |
| "grad_norm": 0.35150127822824473, | |
| "learning_rate": 4.4046732879497295e-05, | |
| "loss": 0.4333, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5696887686062246, | |
| "grad_norm": 0.25191342410729517, | |
| "learning_rate": 4.354231293958801e-05, | |
| "loss": 0.4314, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.5805142083897157, | |
| "grad_norm": 0.3964569105966751, | |
| "learning_rate": 4.3037324569009854e-05, | |
| "loss": 0.4315, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5913396481732072, | |
| "grad_norm": 0.4506273296490119, | |
| "learning_rate": 4.2531848802635264e-05, | |
| "loss": 0.4279, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6021650879566982, | |
| "grad_norm": 0.3792095172740697, | |
| "learning_rate": 4.202596675354851e-05, | |
| "loss": 0.4476, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6129905277401895, | |
| "grad_norm": 0.25609624572619266, | |
| "learning_rate": 4.151975960002958e-05, | |
| "loss": 0.4296, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6238159675236807, | |
| "grad_norm": 0.29780854410214036, | |
| "learning_rate": 4.101330857252752e-05, | |
| "loss": 0.4365, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6346414073071718, | |
| "grad_norm": 0.4435006947829412, | |
| "learning_rate": 4.050669494062561e-05, | |
| "loss": 0.437, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.645466847090663, | |
| "grad_norm": 0.309041123839106, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4305, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6562922868741543, | |
| "grad_norm": 0.2570236992212162, | |
| "learning_rate": 3.9493305059374405e-05, | |
| "loss": 0.4433, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6671177266576453, | |
| "grad_norm": 0.32770704547134333, | |
| "learning_rate": 3.8986691427472496e-05, | |
| "loss": 0.4356, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6779431664411368, | |
| "grad_norm": 0.3077253319134328, | |
| "learning_rate": 3.8480240399970436e-05, | |
| "loss": 0.4287, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6887686062246279, | |
| "grad_norm": 0.25299607930186807, | |
| "learning_rate": 3.7974033246451496e-05, | |
| "loss": 0.4368, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.699594046008119, | |
| "grad_norm": 0.22312926404966466, | |
| "learning_rate": 3.746815119736475e-05, | |
| "loss": 0.4229, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7104194857916104, | |
| "grad_norm": 0.2417526738867689, | |
| "learning_rate": 3.696267543099016e-05, | |
| "loss": 0.4342, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7212449255751014, | |
| "grad_norm": 0.302965747482617, | |
| "learning_rate": 3.6457687060412e-05, | |
| "loss": 0.4408, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7320703653585927, | |
| "grad_norm": 0.20920899613389016, | |
| "learning_rate": 3.595326712050272e-05, | |
| "loss": 0.4347, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.742895805142084, | |
| "grad_norm": 0.25748536345096484, | |
| "learning_rate": 3.5449496554919414e-05, | |
| "loss": 0.4316, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.753721244925575, | |
| "grad_norm": 0.2442961455059911, | |
| "learning_rate": 3.494645620311484e-05, | |
| "loss": 0.4369, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.7645466847090663, | |
| "grad_norm": 0.2193742738754088, | |
| "learning_rate": 3.444422678736525e-05, | |
| "loss": 0.4307, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7753721244925575, | |
| "grad_norm": 0.17715305671292078, | |
| "learning_rate": 3.394288889981695e-05, | |
| "loss": 0.4281, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7861975642760486, | |
| "grad_norm": 0.24502275884699662, | |
| "learning_rate": 3.34425229895538e-05, | |
| "loss": 0.4355, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.79702300405954, | |
| "grad_norm": 0.2192086541292272, | |
| "learning_rate": 3.294320934968768e-05, | |
| "loss": 0.4282, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8078484438430311, | |
| "grad_norm": 0.21694854030399527, | |
| "learning_rate": 3.2445028104473845e-05, | |
| "loss": 0.4338, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8186738836265224, | |
| "grad_norm": 0.2255599010213592, | |
| "learning_rate": 3.194805919645359e-05, | |
| "loss": 0.4306, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8294993234100136, | |
| "grad_norm": 0.16169324703842375, | |
| "learning_rate": 3.145238237362596e-05, | |
| "loss": 0.4386, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8403247631935047, | |
| "grad_norm": 0.19307322773361396, | |
| "learning_rate": 3.0958077176650606e-05, | |
| "loss": 0.4296, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.851150202976996, | |
| "grad_norm": 0.193815929380636, | |
| "learning_rate": 3.0465222926084036e-05, | |
| "loss": 0.4258, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.8619756427604872, | |
| "grad_norm": 0.17043868718230246, | |
| "learning_rate": 2.997389870965118e-05, | |
| "loss": 0.427, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.8728010825439783, | |
| "grad_norm": 0.23106100080622996, | |
| "learning_rate": 2.948418336955421e-05, | |
| "loss": 0.4209, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.8836265223274695, | |
| "grad_norm": 0.19492435702475458, | |
| "learning_rate": 2.899615548982088e-05, | |
| "loss": 0.4335, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.8944519621109608, | |
| "grad_norm": 0.18373241681022587, | |
| "learning_rate": 2.8509893383694213e-05, | |
| "loss": 0.4273, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9052774018944518, | |
| "grad_norm": 0.21935000179277805, | |
| "learning_rate": 2.8025475081065684e-05, | |
| "loss": 0.4226, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9161028416779433, | |
| "grad_norm": 0.2081378352848529, | |
| "learning_rate": 2.754297831595385e-05, | |
| "loss": 0.424, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9269282814614344, | |
| "grad_norm": 0.2191334718271232, | |
| "learning_rate": 2.7062480514030478e-05, | |
| "loss": 0.4339, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9377537212449256, | |
| "grad_norm": 0.24225915258539527, | |
| "learning_rate": 2.658405878019612e-05, | |
| "loss": 0.4279, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.9485791610284169, | |
| "grad_norm": 0.18137087946413424, | |
| "learning_rate": 2.6107789886207195e-05, | |
| "loss": 0.4306, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.959404600811908, | |
| "grad_norm": 0.19279540790814204, | |
| "learning_rate": 2.563375025835652e-05, | |
| "loss": 0.4288, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.9702300405953992, | |
| "grad_norm": 0.18225770391566753, | |
| "learning_rate": 2.5162015965209295e-05, | |
| "loss": 0.4257, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9810554803788905, | |
| "grad_norm": 0.1567588227927359, | |
| "learning_rate": 2.4692662705396412e-05, | |
| "loss": 0.4332, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.9918809201623815, | |
| "grad_norm": 0.18452714103356302, | |
| "learning_rate": 2.4225765795467267e-05, | |
| "loss": 0.434, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.002706359945873, | |
| "grad_norm": 0.3038807377018833, | |
| "learning_rate": 2.376140015780372e-05, | |
| "loss": 0.6821, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.013531799729364, | |
| "grad_norm": 0.28735864762072005, | |
| "learning_rate": 2.3299640308597487e-05, | |
| "loss": 0.4037, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.024357239512855, | |
| "grad_norm": 0.18652870105936184, | |
| "learning_rate": 2.2840560345892518e-05, | |
| "loss": 0.4023, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0351826792963466, | |
| "grad_norm": 0.26262420244447227, | |
| "learning_rate": 2.2384233937694626e-05, | |
| "loss": 0.4028, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.0460081190798376, | |
| "grad_norm": 0.21554665169315776, | |
| "learning_rate": 2.1930734310150116e-05, | |
| "loss": 0.3969, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.0568335588633286, | |
| "grad_norm": 0.2365327578302819, | |
| "learning_rate": 2.1480134235795173e-05, | |
| "loss": 0.4016, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.06765899864682, | |
| "grad_norm": 0.18757983371061362, | |
| "learning_rate": 2.10325060218782e-05, | |
| "loss": 0.4064, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.078484438430311, | |
| "grad_norm": 0.26228528185975086, | |
| "learning_rate": 2.0587921498756768e-05, | |
| "loss": 0.4103, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.089309878213802, | |
| "grad_norm": 0.21513017222473524, | |
| "learning_rate": 2.014645200837105e-05, | |
| "loss": 0.403, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.1001353179972937, | |
| "grad_norm": 0.21886351501773119, | |
| "learning_rate": 1.9708168392795718e-05, | |
| "loss": 0.3949, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.1109607577807847, | |
| "grad_norm": 0.21693742682509493, | |
| "learning_rate": 1.9273140982871936e-05, | |
| "loss": 0.4042, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.121786197564276, | |
| "grad_norm": 0.23546290407137777, | |
| "learning_rate": 1.8841439586921515e-05, | |
| "loss": 0.4029, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.1326116373477673, | |
| "grad_norm": 0.21178864909912662, | |
| "learning_rate": 1.841313347954482e-05, | |
| "loss": 0.398, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.1434370771312583, | |
| "grad_norm": 0.18726986326173004, | |
| "learning_rate": 1.7988291390504348e-05, | |
| "loss": 0.3971, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.15426251691475, | |
| "grad_norm": 0.18271534224017974, | |
| "learning_rate": 1.7566981493695828e-05, | |
| "loss": 0.404, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.165087956698241, | |
| "grad_norm": 0.2147844040709664, | |
| "learning_rate": 1.71492713962083e-05, | |
| "loss": 0.4065, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.175913396481732, | |
| "grad_norm": 0.13272282917253114, | |
| "learning_rate": 1.673522812747544e-05, | |
| "loss": 0.4019, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.1867388362652234, | |
| "grad_norm": 0.1639997741865366, | |
| "learning_rate": 1.6324918128519306e-05, | |
| "loss": 0.3989, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.1975642760487144, | |
| "grad_norm": 0.17720118189615242, | |
| "learning_rate": 1.5918407241288678e-05, | |
| "loss": 0.4055, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.208389715832206, | |
| "grad_norm": 0.138122932939748, | |
| "learning_rate": 1.5515760698093485e-05, | |
| "loss": 0.4039, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.219215155615697, | |
| "grad_norm": 0.17125985703791913, | |
| "learning_rate": 1.511704311113705e-05, | |
| "loss": 0.4016, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.230040595399188, | |
| "grad_norm": 0.1587790419110655, | |
| "learning_rate": 1.4722318462147844e-05, | |
| "loss": 0.4021, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.2408660351826795, | |
| "grad_norm": 0.15781237529891373, | |
| "learning_rate": 1.4331650092112406e-05, | |
| "loss": 0.4065, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.2516914749661705, | |
| "grad_norm": 0.15382743811108052, | |
| "learning_rate": 1.394510069111112e-05, | |
| "loss": 0.3967, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.2625169147496615, | |
| "grad_norm": 0.16072099415521288, | |
| "learning_rate": 1.3562732288258377e-05, | |
| "loss": 0.4022, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.273342354533153, | |
| "grad_norm": 0.1263113164901036, | |
| "learning_rate": 1.3184606241748857e-05, | |
| "loss": 0.3976, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.284167794316644, | |
| "grad_norm": 0.15013101349012392, | |
| "learning_rate": 1.2810783229011486e-05, | |
| "loss": 0.3988, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.2949932341001356, | |
| "grad_norm": 0.14736474180918321, | |
| "learning_rate": 1.2441323236972536e-05, | |
| "loss": 0.3994, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.3058186738836266, | |
| "grad_norm": 0.13153103120475906, | |
| "learning_rate": 1.2076285552429642e-05, | |
| "loss": 0.4068, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.3166441136671176, | |
| "grad_norm": 0.13114498339206726, | |
| "learning_rate": 1.1715728752538103e-05, | |
| "loss": 0.4013, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.3274695534506087, | |
| "grad_norm": 0.13515923306986954, | |
| "learning_rate": 1.1359710695411086e-05, | |
| "loss": 0.4023, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.3382949932341, | |
| "grad_norm": 0.13810939452372453, | |
| "learning_rate": 1.100828851083521e-05, | |
| "loss": 0.408, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.349120433017591, | |
| "grad_norm": 0.13827451287992423, | |
| "learning_rate": 1.0661518591102973e-05, | |
| "loss": 0.4015, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.3599458728010827, | |
| "grad_norm": 0.13316022171250266, | |
| "learning_rate": 1.0319456581963578e-05, | |
| "loss": 0.4094, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.3707713125845737, | |
| "grad_norm": 0.13455083798097392, | |
| "learning_rate": 9.982157373693502e-06, | |
| "loss": 0.4064, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.381596752368065, | |
| "grad_norm": 0.13687639487286918, | |
| "learning_rate": 9.649675092288366e-06, | |
| "loss": 0.3935, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.3924221921515563, | |
| "grad_norm": 0.13576731970561884, | |
| "learning_rate": 9.322063090777331e-06, | |
| "loss": 0.4013, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.4032476319350473, | |
| "grad_norm": 0.12695332075481375, | |
| "learning_rate": 8.99937394066165e-06, | |
| "loss": 0.4074, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.414073071718539, | |
| "grad_norm": 0.12992836433950913, | |
| "learning_rate": 8.681659423478587e-06, | |
| "loss": 0.4068, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.42489851150203, | |
| "grad_norm": 0.13614933415740577, | |
| "learning_rate": 8.368970522492064e-06, | |
| "loss": 0.4038, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.435723951285521, | |
| "grad_norm": 0.1299611166958906, | |
| "learning_rate": 8.06135741451146e-06, | |
| "loss": 0.4018, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.4465493910690124, | |
| "grad_norm": 0.12922732037406734, | |
| "learning_rate": 7.758869461839808e-06, | |
| "loss": 0.4049, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.4573748308525034, | |
| "grad_norm": 0.11409530501821215, | |
| "learning_rate": 7.461555204352655e-06, | |
| "loss": 0.3983, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.4682002706359945, | |
| "grad_norm": 0.13109519380435983, | |
| "learning_rate": 7.169462351708958e-06, | |
| "loss": 0.4004, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.479025710419486, | |
| "grad_norm": 0.1096916737922998, | |
| "learning_rate": 6.882637775695147e-06, | |
| "loss": 0.3968, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.489851150202977, | |
| "grad_norm": 0.10726936604984846, | |
| "learning_rate": 6.60112750270371e-06, | |
| "loss": 0.3982, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.500676589986468, | |
| "grad_norm": 0.11479168890681679, | |
| "learning_rate": 6.324976706347317e-06, | |
| "loss": 0.4073, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.5115020297699595, | |
| "grad_norm": 0.10858787032052403, | |
| "learning_rate": 6.054229700209959e-06, | |
| "loss": 0.4048, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.5223274695534506, | |
| "grad_norm": 0.10800689312775234, | |
| "learning_rate": 5.788929930735916e-06, | |
| "loss": 0.3997, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.533152909336942, | |
| "grad_norm": 0.1015329347217822, | |
| "learning_rate": 5.529119970258014e-06, | |
| "loss": 0.4028, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.543978349120433, | |
| "grad_norm": 0.10284628660982224, | |
| "learning_rate": 5.274841510166062e-06, | |
| "loss": 0.3953, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.554803788903924, | |
| "grad_norm": 0.1011980332411963, | |
| "learning_rate": 5.026135354216717e-06, | |
| "loss": 0.3948, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.565629228687415, | |
| "grad_norm": 0.09800734193393426, | |
| "learning_rate": 4.783041411985716e-06, | |
| "loss": 0.4021, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.5764546684709067, | |
| "grad_norm": 0.09700053551886881, | |
| "learning_rate": 4.545598692463675e-06, | |
| "loss": 0.4007, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.5872801082543977, | |
| "grad_norm": 0.10037984607100636, | |
| "learning_rate": 4.3138452977963266e-06, | |
| "loss": 0.3982, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.598105548037889, | |
| "grad_norm": 0.1014350319561999, | |
| "learning_rate": 4.087818417170337e-06, | |
| "loss": 0.4053, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.60893098782138, | |
| "grad_norm": 0.09850776364380359, | |
| "learning_rate": 3.867554320845601e-06, | |
| "loss": 0.4079, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.6197564276048713, | |
| "grad_norm": 0.09833019554471717, | |
| "learning_rate": 3.6530883543350038e-06, | |
| "loss": 0.4072, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.6305818673883627, | |
| "grad_norm": 0.09477548233477305, | |
| "learning_rate": 3.4444549327325325e-06, | |
| "loss": 0.3981, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.641407307171854, | |
| "grad_norm": 0.10863775134112984, | |
| "learning_rate": 3.241687535190776e-06, | |
| "loss": 0.3965, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.6522327469553453, | |
| "grad_norm": 0.1059231743088096, | |
| "learning_rate": 3.0448186995485307e-06, | |
| "loss": 0.3973, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.6630581867388363, | |
| "grad_norm": 0.0984891746140367, | |
| "learning_rate": 2.853880017109516e-06, | |
| "loss": 0.3981, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.6738836265223274, | |
| "grad_norm": 0.09601668207701894, | |
| "learning_rate": 2.6689021275729366e-06, | |
| "loss": 0.4039, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.6847090663058184, | |
| "grad_norm": 0.10112232940930609, | |
| "learning_rate": 2.489914714116788e-06, | |
| "loss": 0.3992, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.69553450608931, | |
| "grad_norm": 0.10142024172351762, | |
| "learning_rate": 2.316946498634605e-06, | |
| "loss": 0.3952, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.706359945872801, | |
| "grad_norm": 0.0969754986462005, | |
| "learning_rate": 2.1500252371265253e-06, | |
| "loss": 0.4065, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.7171853856562924, | |
| "grad_norm": 0.09689017109227589, | |
| "learning_rate": 1.989177715245307e-06, | |
| "loss": 0.4023, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.7280108254397835, | |
| "grad_norm": 0.08996806789637721, | |
| "learning_rate": 1.8344297439980475e-06, | |
| "loss": 0.4063, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.7388362652232745, | |
| "grad_norm": 0.09222182196227491, | |
| "learning_rate": 1.685806155604346e-06, | |
| "loss": 0.4049, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.749661705006766, | |
| "grad_norm": 0.09210760604624076, | |
| "learning_rate": 1.5433307995115043e-06, | |
| "loss": 0.399, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.760487144790257, | |
| "grad_norm": 0.08811433949347404, | |
| "learning_rate": 1.4070265385674176e-06, | |
| "loss": 0.3946, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.7713125845737485, | |
| "grad_norm": 0.09107211149204227, | |
| "learning_rate": 1.276915245351833e-06, | |
| "loss": 0.3933, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.7821380243572396, | |
| "grad_norm": 0.08673352921830904, | |
| "learning_rate": 1.1530177986664425e-06, | |
| "loss": 0.4071, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.7929634641407306, | |
| "grad_norm": 0.08666404171970574, | |
| "learning_rate": 1.0353540801845229e-06, | |
| "loss": 0.3998, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.803788903924222, | |
| "grad_norm": 0.089018119071375, | |
| "learning_rate": 9.239429712605274e-07, | |
| "loss": 0.3985, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.814614343707713, | |
| "grad_norm": 0.08297666448416004, | |
| "learning_rate": 8.188023499002206e-07, | |
| "loss": 0.4038, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.825439783491204, | |
| "grad_norm": 0.08536059982474015, | |
| "learning_rate": 7.199490878918314e-07, | |
| "loss": 0.3938, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.8362652232746957, | |
| "grad_norm": 0.08291161380222195, | |
| "learning_rate": 6.273990480986314e-07, | |
| "loss": 0.399, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.8470906630581867, | |
| "grad_norm": 0.08470090318649044, | |
| "learning_rate": 5.411670819134651e-07, | |
| "loss": 0.4034, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.8579161028416777, | |
| "grad_norm": 0.08630391724430683, | |
| "learning_rate": 4.6126702687554483e-07, | |
| "loss": 0.3989, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.8687415426251692, | |
| "grad_norm": 0.08359943242372093, | |
| "learning_rate": 3.8771170444996895e-07, | |
| "loss": 0.4029, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.8795669824086603, | |
| "grad_norm": 0.08397753527566773, | |
| "learning_rate": 3.2051291797027925e-07, | |
| "loss": 0.4017, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.8903924221921518, | |
| "grad_norm": 0.08527296501600641, | |
| "learning_rate": 2.59681450744389e-07, | |
| "loss": 0.397, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.901217861975643, | |
| "grad_norm": 0.08082037771426254, | |
| "learning_rate": 2.0522706432419382e-07, | |
| "loss": 0.3986, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.912043301759134, | |
| "grad_norm": 0.08133863221177037, | |
| "learning_rate": 1.5715849693916264e-07, | |
| "loss": 0.4004, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.9228687415426253, | |
| "grad_norm": 0.08213439600809583, | |
| "learning_rate": 1.1548346209410366e-07, | |
| "loss": 0.4017, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.9336941813261164, | |
| "grad_norm": 0.0840670086317821, | |
| "learning_rate": 8.020864733140343e-08, | |
| "loss": 0.3987, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.944519621109608, | |
| "grad_norm": 0.08197353679069497, | |
| "learning_rate": 5.133971315788966e-08, | |
| "loss": 0.4047, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.955345060893099, | |
| "grad_norm": 0.08476922218479478, | |
| "learning_rate": 2.8881292136468952e-08, | |
| "loss": 0.4019, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.96617050067659, | |
| "grad_norm": 0.08609409784588085, | |
| "learning_rate": 1.2836988142779228e-08, | |
| "loss": 0.4006, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.976995940460081, | |
| "grad_norm": 0.0821345909006518, | |
| "learning_rate": 3.209375786856761e-09, | |
| "loss": 0.4089, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "grad_norm": 0.07978409748230415, | |
| "learning_rate": 0.0, | |
| "loss": 0.3962, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.9878213802435725, | |
| "step": 276, | |
| "total_flos": 7.10230108051551e+18, | |
| "train_loss": 0.464179332813491, | |
| "train_runtime": 60729.3117, | |
| "train_samples_per_second": 2.336, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 276, | |
| "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": 7.10230108051551e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |