diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,61255 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 10.0, + "eval_steps": 1908, + "global_step": 38150, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.001310615989515072, + "grad_norm": 78.90444946289062, + "learning_rate": 5.2424639580602885e-08, + "loss": 6.7986, + "num_input_tokens_seen": 1184, + "step": 5 + }, + { + "epoch": 0.002621231979030144, + "grad_norm": 41.2728271484375, + "learning_rate": 1.1795543905635651e-07, + "loss": 6.459, + "num_input_tokens_seen": 2448, + "step": 10 + }, + { + "epoch": 0.003931847968545216, + "grad_norm": 41.36561965942383, + "learning_rate": 1.8348623853211012e-07, + "loss": 6.3342, + "num_input_tokens_seen": 4128, + "step": 15 + }, + { + "epoch": 0.005242463958060288, + "grad_norm": 81.78801727294922, + "learning_rate": 2.490170380078637e-07, + "loss": 6.3733, + "num_input_tokens_seen": 5664, + "step": 20 + }, + { + "epoch": 0.00655307994757536, + "grad_norm": 64.42354583740234, + "learning_rate": 3.1454783748361734e-07, + "loss": 6.993, + "num_input_tokens_seen": 7040, + "step": 25 + }, + { + "epoch": 0.007863695937090432, + "grad_norm": 65.59208679199219, + "learning_rate": 3.8007863695937093e-07, + "loss": 6.2515, + "num_input_tokens_seen": 8352, + "step": 30 + }, + { + "epoch": 0.009174311926605505, + "grad_norm": 44.62382507324219, + "learning_rate": 4.4560943643512453e-07, + "loss": 6.3401, + "num_input_tokens_seen": 9680, + "step": 35 + }, + { + "epoch": 0.010484927916120577, + "grad_norm": 39.547264099121094, + "learning_rate": 5.111402359108782e-07, + "loss": 6.0357, + "num_input_tokens_seen": 11424, + "step": 40 + }, + { + "epoch": 0.011795543905635648, + "grad_norm": 70.14107513427734, + "learning_rate": 5.766710353866317e-07, + "loss": 6.1273, + "num_input_tokens_seen": 12544, + "step": 45 + }, + { + "epoch": 0.01310615989515072, + "grad_norm": 59.80147933959961, + "learning_rate": 6.422018348623854e-07, + "loss": 5.9301, + "num_input_tokens_seen": 14112, + "step": 50 + }, + { + "epoch": 0.014416775884665793, + "grad_norm": 41.51630783081055, + "learning_rate": 7.07732634338139e-07, + "loss": 6.1987, + "num_input_tokens_seen": 15232, + "step": 55 + }, + { + "epoch": 0.015727391874180863, + "grad_norm": 52.977909088134766, + "learning_rate": 7.732634338138926e-07, + "loss": 5.5467, + "num_input_tokens_seen": 16816, + "step": 60 + }, + { + "epoch": 0.01703800786369594, + "grad_norm": 52.400447845458984, + "learning_rate": 8.387942332896462e-07, + "loss": 5.162, + "num_input_tokens_seen": 17760, + "step": 65 + }, + { + "epoch": 0.01834862385321101, + "grad_norm": 47.6863899230957, + "learning_rate": 9.043250327653998e-07, + "loss": 4.9736, + "num_input_tokens_seen": 18976, + "step": 70 + }, + { + "epoch": 0.019659239842726082, + "grad_norm": 43.056419372558594, + "learning_rate": 9.698558322411533e-07, + "loss": 4.9793, + "num_input_tokens_seen": 20368, + "step": 75 + }, + { + "epoch": 0.020969855832241154, + "grad_norm": 42.85586166381836, + "learning_rate": 1.035386631716907e-06, + "loss": 4.6473, + "num_input_tokens_seen": 21456, + "step": 80 + }, + { + "epoch": 0.022280471821756225, + "grad_norm": 66.81767272949219, + "learning_rate": 1.1009174311926608e-06, + "loss": 4.7327, + "num_input_tokens_seen": 22592, + "step": 85 + }, + { + "epoch": 0.023591087811271297, + "grad_norm": 35.31460189819336, + "learning_rate": 1.1664482306684142e-06, + "loss": 4.7761, + "num_input_tokens_seen": 23728, + "step": 90 + }, + { + "epoch": 0.02490170380078637, + "grad_norm": 39.3112678527832, + "learning_rate": 1.2319790301441677e-06, + "loss": 4.1649, + "num_input_tokens_seen": 24864, + "step": 95 + }, + { + "epoch": 0.02621231979030144, + "grad_norm": 32.51930236816406, + "learning_rate": 1.2975098296199214e-06, + "loss": 4.3159, + "num_input_tokens_seen": 26272, + "step": 100 + }, + { + "epoch": 0.027522935779816515, + "grad_norm": 31.23662757873535, + "learning_rate": 1.363040629095675e-06, + "loss": 4.3612, + "num_input_tokens_seen": 27248, + "step": 105 + }, + { + "epoch": 0.028833551769331587, + "grad_norm": 29.54172706604004, + "learning_rate": 1.4285714285714286e-06, + "loss": 3.3342, + "num_input_tokens_seen": 28688, + "step": 110 + }, + { + "epoch": 0.03014416775884666, + "grad_norm": 24.814367294311523, + "learning_rate": 1.4941022280471821e-06, + "loss": 4.3775, + "num_input_tokens_seen": 29792, + "step": 115 + }, + { + "epoch": 0.03145478374836173, + "grad_norm": 24.28786277770996, + "learning_rate": 1.559633027522936e-06, + "loss": 3.5152, + "num_input_tokens_seen": 31424, + "step": 120 + }, + { + "epoch": 0.0327653997378768, + "grad_norm": 36.81103515625, + "learning_rate": 1.6251638269986893e-06, + "loss": 3.867, + "num_input_tokens_seen": 32384, + "step": 125 + }, + { + "epoch": 0.03407601572739188, + "grad_norm": 33.66696548461914, + "learning_rate": 1.690694626474443e-06, + "loss": 3.8969, + "num_input_tokens_seen": 33808, + "step": 130 + }, + { + "epoch": 0.035386631716906945, + "grad_norm": 55.85234069824219, + "learning_rate": 1.7562254259501965e-06, + "loss": 3.6425, + "num_input_tokens_seen": 34912, + "step": 135 + }, + { + "epoch": 0.03669724770642202, + "grad_norm": 20.649677276611328, + "learning_rate": 1.8217562254259502e-06, + "loss": 3.5186, + "num_input_tokens_seen": 36576, + "step": 140 + }, + { + "epoch": 0.03800786369593709, + "grad_norm": 35.337162017822266, + "learning_rate": 1.8872870249017041e-06, + "loss": 3.6483, + "num_input_tokens_seen": 37664, + "step": 145 + }, + { + "epoch": 0.039318479685452164, + "grad_norm": 45.230010986328125, + "learning_rate": 1.9528178243774574e-06, + "loss": 3.6386, + "num_input_tokens_seen": 38816, + "step": 150 + }, + { + "epoch": 0.04062909567496723, + "grad_norm": 32.661529541015625, + "learning_rate": 2.0183486238532113e-06, + "loss": 3.2565, + "num_input_tokens_seen": 40272, + "step": 155 + }, + { + "epoch": 0.04193971166448231, + "grad_norm": 53.766380310058594, + "learning_rate": 2.083879423328965e-06, + "loss": 3.3967, + "num_input_tokens_seen": 41392, + "step": 160 + }, + { + "epoch": 0.04325032765399738, + "grad_norm": 22.959558486938477, + "learning_rate": 2.1494102228047183e-06, + "loss": 3.648, + "num_input_tokens_seen": 43152, + "step": 165 + }, + { + "epoch": 0.04456094364351245, + "grad_norm": 18.39083480834961, + "learning_rate": 2.2149410222804718e-06, + "loss": 3.3635, + "num_input_tokens_seen": 44992, + "step": 170 + }, + { + "epoch": 0.045871559633027525, + "grad_norm": 33.41960144042969, + "learning_rate": 2.2804718217562257e-06, + "loss": 3.0703, + "num_input_tokens_seen": 46192, + "step": 175 + }, + { + "epoch": 0.047182175622542594, + "grad_norm": 52.294857025146484, + "learning_rate": 2.346002621231979e-06, + "loss": 3.4704, + "num_input_tokens_seen": 47216, + "step": 180 + }, + { + "epoch": 0.04849279161205767, + "grad_norm": 27.183841705322266, + "learning_rate": 2.4115334207077327e-06, + "loss": 2.6495, + "num_input_tokens_seen": 48944, + "step": 185 + }, + { + "epoch": 0.04980340760157274, + "grad_norm": 30.92019271850586, + "learning_rate": 2.4770642201834866e-06, + "loss": 2.7644, + "num_input_tokens_seen": 50288, + "step": 190 + }, + { + "epoch": 0.05111402359108781, + "grad_norm": 24.655986785888672, + "learning_rate": 2.54259501965924e-06, + "loss": 2.5845, + "num_input_tokens_seen": 51552, + "step": 195 + }, + { + "epoch": 0.05242463958060288, + "grad_norm": 26.806255340576172, + "learning_rate": 2.6081258191349936e-06, + "loss": 3.2747, + "num_input_tokens_seen": 53328, + "step": 200 + }, + { + "epoch": 0.053735255570117955, + "grad_norm": 33.089813232421875, + "learning_rate": 2.673656618610747e-06, + "loss": 2.5992, + "num_input_tokens_seen": 54640, + "step": 205 + }, + { + "epoch": 0.05504587155963303, + "grad_norm": 32.14365005493164, + "learning_rate": 2.739187418086501e-06, + "loss": 2.7476, + "num_input_tokens_seen": 55776, + "step": 210 + }, + { + "epoch": 0.0563564875491481, + "grad_norm": 18.640207290649414, + "learning_rate": 2.8047182175622545e-06, + "loss": 2.5949, + "num_input_tokens_seen": 57216, + "step": 215 + }, + { + "epoch": 0.057667103538663174, + "grad_norm": 12.660541534423828, + "learning_rate": 2.870249017038008e-06, + "loss": 2.7169, + "num_input_tokens_seen": 59360, + "step": 220 + }, + { + "epoch": 0.05897771952817824, + "grad_norm": 26.890460968017578, + "learning_rate": 2.935779816513762e-06, + "loss": 2.8786, + "num_input_tokens_seen": 60384, + "step": 225 + }, + { + "epoch": 0.06028833551769332, + "grad_norm": 31.395877838134766, + "learning_rate": 3.0013106159895154e-06, + "loss": 2.463, + "num_input_tokens_seen": 61888, + "step": 230 + }, + { + "epoch": 0.061598951507208385, + "grad_norm": 41.00429916381836, + "learning_rate": 3.066841415465269e-06, + "loss": 2.2599, + "num_input_tokens_seen": 62800, + "step": 235 + }, + { + "epoch": 0.06290956749672345, + "grad_norm": 24.173397064208984, + "learning_rate": 3.1323722149410228e-06, + "loss": 2.3209, + "num_input_tokens_seen": 64256, + "step": 240 + }, + { + "epoch": 0.06422018348623854, + "grad_norm": 35.1030158996582, + "learning_rate": 3.1979030144167763e-06, + "loss": 2.4618, + "num_input_tokens_seen": 65728, + "step": 245 + }, + { + "epoch": 0.0655307994757536, + "grad_norm": 18.51744842529297, + "learning_rate": 3.2634338138925293e-06, + "loss": 2.6133, + "num_input_tokens_seen": 66912, + "step": 250 + }, + { + "epoch": 0.06684141546526867, + "grad_norm": 26.241809844970703, + "learning_rate": 3.328964613368283e-06, + "loss": 2.2286, + "num_input_tokens_seen": 67904, + "step": 255 + }, + { + "epoch": 0.06815203145478375, + "grad_norm": 22.378995895385742, + "learning_rate": 3.394495412844037e-06, + "loss": 2.3407, + "num_input_tokens_seen": 69456, + "step": 260 + }, + { + "epoch": 0.06946264744429882, + "grad_norm": 21.67023277282715, + "learning_rate": 3.4600262123197906e-06, + "loss": 2.8786, + "num_input_tokens_seen": 70832, + "step": 265 + }, + { + "epoch": 0.07077326343381389, + "grad_norm": 19.290449142456055, + "learning_rate": 3.5255570117955437e-06, + "loss": 2.7456, + "num_input_tokens_seen": 72448, + "step": 270 + }, + { + "epoch": 0.07208387942332896, + "grad_norm": 23.92256736755371, + "learning_rate": 3.591087811271298e-06, + "loss": 2.5068, + "num_input_tokens_seen": 73920, + "step": 275 + }, + { + "epoch": 0.07339449541284404, + "grad_norm": 54.279457092285156, + "learning_rate": 3.6566186107470515e-06, + "loss": 2.5438, + "num_input_tokens_seen": 75472, + "step": 280 + }, + { + "epoch": 0.07470511140235911, + "grad_norm": 15.394092559814453, + "learning_rate": 3.7221494102228046e-06, + "loss": 2.4037, + "num_input_tokens_seen": 76976, + "step": 285 + }, + { + "epoch": 0.07601572739187418, + "grad_norm": 21.600587844848633, + "learning_rate": 3.787680209698558e-06, + "loss": 2.3192, + "num_input_tokens_seen": 78496, + "step": 290 + }, + { + "epoch": 0.07732634338138926, + "grad_norm": 26.971454620361328, + "learning_rate": 3.853211009174312e-06, + "loss": 1.9689, + "num_input_tokens_seen": 80032, + "step": 295 + }, + { + "epoch": 0.07863695937090433, + "grad_norm": 23.49415397644043, + "learning_rate": 3.918741808650066e-06, + "loss": 1.6909, + "num_input_tokens_seen": 81104, + "step": 300 + }, + { + "epoch": 0.0799475753604194, + "grad_norm": 40.39797592163086, + "learning_rate": 3.984272608125819e-06, + "loss": 1.7138, + "num_input_tokens_seen": 82224, + "step": 305 + }, + { + "epoch": 0.08125819134993446, + "grad_norm": 27.454008102416992, + "learning_rate": 4.049803407601573e-06, + "loss": 1.7154, + "num_input_tokens_seen": 84160, + "step": 310 + }, + { + "epoch": 0.08256880733944955, + "grad_norm": 22.636173248291016, + "learning_rate": 4.115334207077327e-06, + "loss": 2.0075, + "num_input_tokens_seen": 85488, + "step": 315 + }, + { + "epoch": 0.08387942332896461, + "grad_norm": 19.532976150512695, + "learning_rate": 4.18086500655308e-06, + "loss": 2.203, + "num_input_tokens_seen": 86768, + "step": 320 + }, + { + "epoch": 0.08519003931847968, + "grad_norm": 19.757543563842773, + "learning_rate": 4.246395806028834e-06, + "loss": 1.6721, + "num_input_tokens_seen": 87920, + "step": 325 + }, + { + "epoch": 0.08650065530799476, + "grad_norm": 22.11702537536621, + "learning_rate": 4.311926605504588e-06, + "loss": 2.5713, + "num_input_tokens_seen": 89136, + "step": 330 + }, + { + "epoch": 0.08781127129750983, + "grad_norm": 18.782543182373047, + "learning_rate": 4.377457404980341e-06, + "loss": 1.9157, + "num_input_tokens_seen": 90464, + "step": 335 + }, + { + "epoch": 0.0891218872870249, + "grad_norm": 17.234968185424805, + "learning_rate": 4.442988204456095e-06, + "loss": 1.8514, + "num_input_tokens_seen": 91984, + "step": 340 + }, + { + "epoch": 0.09043250327653997, + "grad_norm": 25.50652313232422, + "learning_rate": 4.508519003931848e-06, + "loss": 1.3482, + "num_input_tokens_seen": 93024, + "step": 345 + }, + { + "epoch": 0.09174311926605505, + "grad_norm": 47.17985153198242, + "learning_rate": 4.574049803407602e-06, + "loss": 1.499, + "num_input_tokens_seen": 94464, + "step": 350 + }, + { + "epoch": 0.09305373525557012, + "grad_norm": 19.23650360107422, + "learning_rate": 4.639580602883356e-06, + "loss": 1.7817, + "num_input_tokens_seen": 95664, + "step": 355 + }, + { + "epoch": 0.09436435124508519, + "grad_norm": 52.37240219116211, + "learning_rate": 4.705111402359109e-06, + "loss": 1.4435, + "num_input_tokens_seen": 96608, + "step": 360 + }, + { + "epoch": 0.09567496723460026, + "grad_norm": 12.585490226745605, + "learning_rate": 4.7706422018348626e-06, + "loss": 1.3371, + "num_input_tokens_seen": 97968, + "step": 365 + }, + { + "epoch": 0.09698558322411534, + "grad_norm": 20.521448135375977, + "learning_rate": 4.8361730013106165e-06, + "loss": 1.185, + "num_input_tokens_seen": 99152, + "step": 370 + }, + { + "epoch": 0.0982961992136304, + "grad_norm": 18.46451759338379, + "learning_rate": 4.9017038007863695e-06, + "loss": 1.2584, + "num_input_tokens_seen": 100592, + "step": 375 + }, + { + "epoch": 0.09960681520314547, + "grad_norm": 18.49262237548828, + "learning_rate": 4.9672346002621235e-06, + "loss": 1.0506, + "num_input_tokens_seen": 101856, + "step": 380 + }, + { + "epoch": 0.10091743119266056, + "grad_norm": 9.087496757507324, + "learning_rate": 5.032765399737877e-06, + "loss": 1.079, + "num_input_tokens_seen": 103024, + "step": 385 + }, + { + "epoch": 0.10222804718217562, + "grad_norm": 20.173110961914062, + "learning_rate": 5.0982961992136304e-06, + "loss": 0.7351, + "num_input_tokens_seen": 104512, + "step": 390 + }, + { + "epoch": 0.10353866317169069, + "grad_norm": 13.125005722045898, + "learning_rate": 5.163826998689384e-06, + "loss": 0.7511, + "num_input_tokens_seen": 105904, + "step": 395 + }, + { + "epoch": 0.10484927916120576, + "grad_norm": 12.73207950592041, + "learning_rate": 5.229357798165138e-06, + "loss": 0.8194, + "num_input_tokens_seen": 107088, + "step": 400 + }, + { + "epoch": 0.10615989515072084, + "grad_norm": 10.943971633911133, + "learning_rate": 5.294888597640891e-06, + "loss": 1.0875, + "num_input_tokens_seen": 108240, + "step": 405 + }, + { + "epoch": 0.10747051114023591, + "grad_norm": 6.830711841583252, + "learning_rate": 5.360419397116645e-06, + "loss": 0.818, + "num_input_tokens_seen": 109472, + "step": 410 + }, + { + "epoch": 0.10878112712975098, + "grad_norm": 11.10379409790039, + "learning_rate": 5.425950196592398e-06, + "loss": 0.8605, + "num_input_tokens_seen": 111056, + "step": 415 + }, + { + "epoch": 0.11009174311926606, + "grad_norm": 10.04479694366455, + "learning_rate": 5.491480996068152e-06, + "loss": 0.6954, + "num_input_tokens_seen": 112448, + "step": 420 + }, + { + "epoch": 0.11140235910878113, + "grad_norm": 10.219340324401855, + "learning_rate": 5.557011795543906e-06, + "loss": 0.8871, + "num_input_tokens_seen": 113648, + "step": 425 + }, + { + "epoch": 0.1127129750982962, + "grad_norm": 6.6353535652160645, + "learning_rate": 5.622542595019659e-06, + "loss": 1.2326, + "num_input_tokens_seen": 114864, + "step": 430 + }, + { + "epoch": 0.11402359108781127, + "grad_norm": 18.15001678466797, + "learning_rate": 5.688073394495413e-06, + "loss": 0.8113, + "num_input_tokens_seen": 115968, + "step": 435 + }, + { + "epoch": 0.11533420707732635, + "grad_norm": 8.346098899841309, + "learning_rate": 5.753604193971167e-06, + "loss": 0.8079, + "num_input_tokens_seen": 117680, + "step": 440 + }, + { + "epoch": 0.11664482306684142, + "grad_norm": 7.718857765197754, + "learning_rate": 5.81913499344692e-06, + "loss": 0.7636, + "num_input_tokens_seen": 118720, + "step": 445 + }, + { + "epoch": 0.11795543905635648, + "grad_norm": 5.09888219833374, + "learning_rate": 5.884665792922674e-06, + "loss": 0.5719, + "num_input_tokens_seen": 120288, + "step": 450 + }, + { + "epoch": 0.11926605504587157, + "grad_norm": 54.135658264160156, + "learning_rate": 5.950196592398428e-06, + "loss": 0.701, + "num_input_tokens_seen": 121728, + "step": 455 + }, + { + "epoch": 0.12057667103538663, + "grad_norm": 8.577695846557617, + "learning_rate": 6.015727391874181e-06, + "loss": 0.6586, + "num_input_tokens_seen": 122880, + "step": 460 + }, + { + "epoch": 0.1218872870249017, + "grad_norm": 11.658777236938477, + "learning_rate": 6.081258191349935e-06, + "loss": 0.7658, + "num_input_tokens_seen": 124112, + "step": 465 + }, + { + "epoch": 0.12319790301441677, + "grad_norm": 12.62354564666748, + "learning_rate": 6.146788990825689e-06, + "loss": 0.954, + "num_input_tokens_seen": 125568, + "step": 470 + }, + { + "epoch": 0.12450851900393185, + "grad_norm": 6.038365364074707, + "learning_rate": 6.212319790301442e-06, + "loss": 0.6385, + "num_input_tokens_seen": 127216, + "step": 475 + }, + { + "epoch": 0.1258191349934469, + "grad_norm": 10.674358367919922, + "learning_rate": 6.277850589777196e-06, + "loss": 0.9186, + "num_input_tokens_seen": 128560, + "step": 480 + }, + { + "epoch": 0.127129750982962, + "grad_norm": 9.93392562866211, + "learning_rate": 6.343381389252949e-06, + "loss": 0.6429, + "num_input_tokens_seen": 129952, + "step": 485 + }, + { + "epoch": 0.12844036697247707, + "grad_norm": 6.382375240325928, + "learning_rate": 6.408912188728703e-06, + "loss": 0.8343, + "num_input_tokens_seen": 130976, + "step": 490 + }, + { + "epoch": 0.12975098296199214, + "grad_norm": 8.834505081176758, + "learning_rate": 6.474442988204456e-06, + "loss": 0.4068, + "num_input_tokens_seen": 132256, + "step": 495 + }, + { + "epoch": 0.1310615989515072, + "grad_norm": 7.0455403327941895, + "learning_rate": 6.539973787680211e-06, + "loss": 0.8403, + "num_input_tokens_seen": 134256, + "step": 500 + }, + { + "epoch": 0.13237221494102228, + "grad_norm": 4.3123860359191895, + "learning_rate": 6.605504587155964e-06, + "loss": 0.5071, + "num_input_tokens_seen": 135584, + "step": 505 + }, + { + "epoch": 0.13368283093053734, + "grad_norm": 7.081425666809082, + "learning_rate": 6.671035386631718e-06, + "loss": 0.6415, + "num_input_tokens_seen": 136576, + "step": 510 + }, + { + "epoch": 0.1349934469200524, + "grad_norm": 10.290603637695312, + "learning_rate": 6.736566186107471e-06, + "loss": 0.5982, + "num_input_tokens_seen": 137952, + "step": 515 + }, + { + "epoch": 0.1363040629095675, + "grad_norm": 6.198872089385986, + "learning_rate": 6.8020969855832246e-06, + "loss": 0.5363, + "num_input_tokens_seen": 139024, + "step": 520 + }, + { + "epoch": 0.13761467889908258, + "grad_norm": 9.0060396194458, + "learning_rate": 6.867627785058978e-06, + "loss": 0.5653, + "num_input_tokens_seen": 140464, + "step": 525 + }, + { + "epoch": 0.13892529488859764, + "grad_norm": 28.327672958374023, + "learning_rate": 6.933158584534731e-06, + "loss": 0.5365, + "num_input_tokens_seen": 141824, + "step": 530 + }, + { + "epoch": 0.1402359108781127, + "grad_norm": 11.991899490356445, + "learning_rate": 6.9986893840104855e-06, + "loss": 0.5991, + "num_input_tokens_seen": 142864, + "step": 535 + }, + { + "epoch": 0.14154652686762778, + "grad_norm": 8.132561683654785, + "learning_rate": 7.064220183486239e-06, + "loss": 0.8474, + "num_input_tokens_seen": 143856, + "step": 540 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 6.103424549102783, + "learning_rate": 7.1297509829619924e-06, + "loss": 0.5345, + "num_input_tokens_seen": 145248, + "step": 545 + }, + { + "epoch": 0.14416775884665792, + "grad_norm": 3.0450127124786377, + "learning_rate": 7.195281782437746e-06, + "loss": 0.5313, + "num_input_tokens_seen": 146912, + "step": 550 + }, + { + "epoch": 0.145478374836173, + "grad_norm": 7.714560031890869, + "learning_rate": 7.260812581913499e-06, + "loss": 0.5775, + "num_input_tokens_seen": 148224, + "step": 555 + }, + { + "epoch": 0.14678899082568808, + "grad_norm": 5.751674652099609, + "learning_rate": 7.326343381389253e-06, + "loss": 0.5419, + "num_input_tokens_seen": 149568, + "step": 560 + }, + { + "epoch": 0.14809960681520315, + "grad_norm": 11.17633056640625, + "learning_rate": 7.391874180865006e-06, + "loss": 0.6886, + "num_input_tokens_seen": 150912, + "step": 565 + }, + { + "epoch": 0.14941022280471822, + "grad_norm": 5.151185512542725, + "learning_rate": 7.457404980340761e-06, + "loss": 0.6381, + "num_input_tokens_seen": 152352, + "step": 570 + }, + { + "epoch": 0.15072083879423329, + "grad_norm": 4.503359317779541, + "learning_rate": 7.522935779816514e-06, + "loss": 0.4535, + "num_input_tokens_seen": 153728, + "step": 575 + }, + { + "epoch": 0.15203145478374835, + "grad_norm": 40.414730072021484, + "learning_rate": 7.588466579292268e-06, + "loss": 1.0023, + "num_input_tokens_seen": 155200, + "step": 580 + }, + { + "epoch": 0.15334207077326342, + "grad_norm": 6.218530178070068, + "learning_rate": 7.653997378768021e-06, + "loss": 0.6087, + "num_input_tokens_seen": 156288, + "step": 585 + }, + { + "epoch": 0.15465268676277852, + "grad_norm": 17.098793029785156, + "learning_rate": 7.719528178243775e-06, + "loss": 0.6053, + "num_input_tokens_seen": 157664, + "step": 590 + }, + { + "epoch": 0.1559633027522936, + "grad_norm": 6.373294830322266, + "learning_rate": 7.785058977719529e-06, + "loss": 0.3845, + "num_input_tokens_seen": 158832, + "step": 595 + }, + { + "epoch": 0.15727391874180865, + "grad_norm": 16.252782821655273, + "learning_rate": 7.850589777195281e-06, + "loss": 0.5163, + "num_input_tokens_seen": 159984, + "step": 600 + }, + { + "epoch": 0.15858453473132372, + "grad_norm": 7.853078365325928, + "learning_rate": 7.916120576671037e-06, + "loss": 0.5095, + "num_input_tokens_seen": 161248, + "step": 605 + }, + { + "epoch": 0.1598951507208388, + "grad_norm": 6.31785249710083, + "learning_rate": 7.981651376146789e-06, + "loss": 0.6285, + "num_input_tokens_seen": 162608, + "step": 610 + }, + { + "epoch": 0.16120576671035386, + "grad_norm": 8.425243377685547, + "learning_rate": 8.047182175622543e-06, + "loss": 0.5152, + "num_input_tokens_seen": 163696, + "step": 615 + }, + { + "epoch": 0.16251638269986893, + "grad_norm": 5.220122337341309, + "learning_rate": 8.112712975098297e-06, + "loss": 0.6275, + "num_input_tokens_seen": 165056, + "step": 620 + }, + { + "epoch": 0.16382699868938402, + "grad_norm": 5.324240207672119, + "learning_rate": 8.17824377457405e-06, + "loss": 0.4361, + "num_input_tokens_seen": 166848, + "step": 625 + }, + { + "epoch": 0.1651376146788991, + "grad_norm": 10.30737590789795, + "learning_rate": 8.243774574049803e-06, + "loss": 0.6254, + "num_input_tokens_seen": 168240, + "step": 630 + }, + { + "epoch": 0.16644823066841416, + "grad_norm": 7.064831256866455, + "learning_rate": 8.309305373525557e-06, + "loss": 0.7787, + "num_input_tokens_seen": 169840, + "step": 635 + }, + { + "epoch": 0.16775884665792923, + "grad_norm": 7.249665260314941, + "learning_rate": 8.374836173001311e-06, + "loss": 0.6294, + "num_input_tokens_seen": 171152, + "step": 640 + }, + { + "epoch": 0.1690694626474443, + "grad_norm": 5.269425392150879, + "learning_rate": 8.440366972477065e-06, + "loss": 0.7506, + "num_input_tokens_seen": 172144, + "step": 645 + }, + { + "epoch": 0.17038007863695936, + "grad_norm": 6.5213704109191895, + "learning_rate": 8.505897771952819e-06, + "loss": 0.5968, + "num_input_tokens_seen": 173440, + "step": 650 + }, + { + "epoch": 0.17169069462647443, + "grad_norm": 28.168981552124023, + "learning_rate": 8.571428571428573e-06, + "loss": 0.6292, + "num_input_tokens_seen": 174928, + "step": 655 + }, + { + "epoch": 0.17300131061598953, + "grad_norm": 3.0344350337982178, + "learning_rate": 8.636959370904325e-06, + "loss": 0.5797, + "num_input_tokens_seen": 176672, + "step": 660 + }, + { + "epoch": 0.1743119266055046, + "grad_norm": 56.66756057739258, + "learning_rate": 8.702490170380079e-06, + "loss": 0.883, + "num_input_tokens_seen": 177712, + "step": 665 + }, + { + "epoch": 0.17562254259501967, + "grad_norm": 8.087437629699707, + "learning_rate": 8.768020969855833e-06, + "loss": 0.773, + "num_input_tokens_seen": 179232, + "step": 670 + }, + { + "epoch": 0.17693315858453473, + "grad_norm": 8.208806991577148, + "learning_rate": 8.833551769331587e-06, + "loss": 0.516, + "num_input_tokens_seen": 180688, + "step": 675 + }, + { + "epoch": 0.1782437745740498, + "grad_norm": 4.524869918823242, + "learning_rate": 8.89908256880734e-06, + "loss": 0.6025, + "num_input_tokens_seen": 181824, + "step": 680 + }, + { + "epoch": 0.17955439056356487, + "grad_norm": 5.313582420349121, + "learning_rate": 8.964613368283094e-06, + "loss": 0.7081, + "num_input_tokens_seen": 182768, + "step": 685 + }, + { + "epoch": 0.18086500655307994, + "grad_norm": 5.022678852081299, + "learning_rate": 9.030144167758847e-06, + "loss": 0.7526, + "num_input_tokens_seen": 184240, + "step": 690 + }, + { + "epoch": 0.182175622542595, + "grad_norm": 12.518412590026855, + "learning_rate": 9.0956749672346e-06, + "loss": 0.5776, + "num_input_tokens_seen": 185200, + "step": 695 + }, + { + "epoch": 0.1834862385321101, + "grad_norm": 7.289027214050293, + "learning_rate": 9.161205766710354e-06, + "loss": 0.4367, + "num_input_tokens_seen": 186464, + "step": 700 + }, + { + "epoch": 0.18479685452162517, + "grad_norm": 3.9767708778381348, + "learning_rate": 9.226736566186107e-06, + "loss": 0.4508, + "num_input_tokens_seen": 187728, + "step": 705 + }, + { + "epoch": 0.18610747051114024, + "grad_norm": 7.315395832061768, + "learning_rate": 9.29226736566186e-06, + "loss": 0.5393, + "num_input_tokens_seen": 188944, + "step": 710 + }, + { + "epoch": 0.1874180865006553, + "grad_norm": 4.227368354797363, + "learning_rate": 9.357798165137616e-06, + "loss": 0.641, + "num_input_tokens_seen": 190288, + "step": 715 + }, + { + "epoch": 0.18872870249017037, + "grad_norm": 8.347152709960938, + "learning_rate": 9.423328964613368e-06, + "loss": 0.398, + "num_input_tokens_seen": 191264, + "step": 720 + }, + { + "epoch": 0.19003931847968544, + "grad_norm": 29.949939727783203, + "learning_rate": 9.488859764089122e-06, + "loss": 0.6598, + "num_input_tokens_seen": 192336, + "step": 725 + }, + { + "epoch": 0.1913499344692005, + "grad_norm": 4.6679253578186035, + "learning_rate": 9.554390563564876e-06, + "loss": 0.5838, + "num_input_tokens_seen": 193584, + "step": 730 + }, + { + "epoch": 0.1926605504587156, + "grad_norm": 6.97373628616333, + "learning_rate": 9.619921363040628e-06, + "loss": 0.5564, + "num_input_tokens_seen": 194800, + "step": 735 + }, + { + "epoch": 0.19397116644823068, + "grad_norm": 13.200016021728516, + "learning_rate": 9.685452162516382e-06, + "loss": 0.5241, + "num_input_tokens_seen": 196192, + "step": 740 + }, + { + "epoch": 0.19528178243774574, + "grad_norm": 19.456085205078125, + "learning_rate": 9.750982961992136e-06, + "loss": 0.5093, + "num_input_tokens_seen": 197200, + "step": 745 + }, + { + "epoch": 0.1965923984272608, + "grad_norm": 6.269876956939697, + "learning_rate": 9.81651376146789e-06, + "loss": 0.7171, + "num_input_tokens_seen": 198384, + "step": 750 + }, + { + "epoch": 0.19790301441677588, + "grad_norm": 3.136441230773926, + "learning_rate": 9.882044560943644e-06, + "loss": 0.6555, + "num_input_tokens_seen": 199680, + "step": 755 + }, + { + "epoch": 0.19921363040629095, + "grad_norm": 4.2629313468933105, + "learning_rate": 9.947575360419398e-06, + "loss": 0.527, + "num_input_tokens_seen": 200960, + "step": 760 + }, + { + "epoch": 0.20052424639580602, + "grad_norm": 6.283478260040283, + "learning_rate": 1.0013106159895152e-05, + "loss": 0.6909, + "num_input_tokens_seen": 203120, + "step": 765 + }, + { + "epoch": 0.2018348623853211, + "grad_norm": 5.0882182121276855, + "learning_rate": 1.0078636959370904e-05, + "loss": 0.798, + "num_input_tokens_seen": 204848, + "step": 770 + }, + { + "epoch": 0.20314547837483618, + "grad_norm": 8.467244148254395, + "learning_rate": 1.0144167758846658e-05, + "loss": 0.6908, + "num_input_tokens_seen": 206080, + "step": 775 + }, + { + "epoch": 0.20445609436435125, + "grad_norm": 3.120208740234375, + "learning_rate": 1.0209698558322412e-05, + "loss": 0.6829, + "num_input_tokens_seen": 207920, + "step": 780 + }, + { + "epoch": 0.20576671035386632, + "grad_norm": 3.7655069828033447, + "learning_rate": 1.0275229357798166e-05, + "loss": 0.6658, + "num_input_tokens_seen": 209040, + "step": 785 + }, + { + "epoch": 0.20707732634338138, + "grad_norm": 2.013965368270874, + "learning_rate": 1.034076015727392e-05, + "loss": 0.5424, + "num_input_tokens_seen": 210384, + "step": 790 + }, + { + "epoch": 0.20838794233289645, + "grad_norm": 4.567400932312012, + "learning_rate": 1.0406290956749674e-05, + "loss": 0.6535, + "num_input_tokens_seen": 212032, + "step": 795 + }, + { + "epoch": 0.20969855832241152, + "grad_norm": 40.72431945800781, + "learning_rate": 1.0471821756225426e-05, + "loss": 0.6656, + "num_input_tokens_seen": 212896, + "step": 800 + }, + { + "epoch": 0.21100917431192662, + "grad_norm": 30.868864059448242, + "learning_rate": 1.053735255570118e-05, + "loss": 0.8836, + "num_input_tokens_seen": 214352, + "step": 805 + }, + { + "epoch": 0.21231979030144169, + "grad_norm": 10.82632064819336, + "learning_rate": 1.0602883355176934e-05, + "loss": 0.7186, + "num_input_tokens_seen": 215664, + "step": 810 + }, + { + "epoch": 0.21363040629095675, + "grad_norm": 4.996654987335205, + "learning_rate": 1.0668414154652686e-05, + "loss": 0.5419, + "num_input_tokens_seen": 217104, + "step": 815 + }, + { + "epoch": 0.21494102228047182, + "grad_norm": 9.035186767578125, + "learning_rate": 1.0733944954128442e-05, + "loss": 0.6325, + "num_input_tokens_seen": 218224, + "step": 820 + }, + { + "epoch": 0.2162516382699869, + "grad_norm": 5.863041400909424, + "learning_rate": 1.0799475753604196e-05, + "loss": 0.5473, + "num_input_tokens_seen": 219584, + "step": 825 + }, + { + "epoch": 0.21756225425950196, + "grad_norm": 3.0107197761535645, + "learning_rate": 1.0865006553079948e-05, + "loss": 0.6417, + "num_input_tokens_seen": 220912, + "step": 830 + }, + { + "epoch": 0.21887287024901703, + "grad_norm": 10.470301628112793, + "learning_rate": 1.0930537352555702e-05, + "loss": 0.4998, + "num_input_tokens_seen": 222272, + "step": 835 + }, + { + "epoch": 0.22018348623853212, + "grad_norm": 2.911963939666748, + "learning_rate": 1.0996068152031456e-05, + "loss": 0.6581, + "num_input_tokens_seen": 223600, + "step": 840 + }, + { + "epoch": 0.2214941022280472, + "grad_norm": 3.2466766834259033, + "learning_rate": 1.1061598951507208e-05, + "loss": 0.8038, + "num_input_tokens_seen": 225232, + "step": 845 + }, + { + "epoch": 0.22280471821756226, + "grad_norm": 11.964326858520508, + "learning_rate": 1.1127129750982962e-05, + "loss": 0.5353, + "num_input_tokens_seen": 226512, + "step": 850 + }, + { + "epoch": 0.22411533420707733, + "grad_norm": 9.433073043823242, + "learning_rate": 1.1192660550458717e-05, + "loss": 0.7659, + "num_input_tokens_seen": 227552, + "step": 855 + }, + { + "epoch": 0.2254259501965924, + "grad_norm": 6.268906116485596, + "learning_rate": 1.125819134993447e-05, + "loss": 0.6396, + "num_input_tokens_seen": 228608, + "step": 860 + }, + { + "epoch": 0.22673656618610746, + "grad_norm": 1.048985242843628, + "learning_rate": 1.1323722149410223e-05, + "loss": 0.369, + "num_input_tokens_seen": 230192, + "step": 865 + }, + { + "epoch": 0.22804718217562253, + "grad_norm": 2.4074795246124268, + "learning_rate": 1.1389252948885977e-05, + "loss": 0.5466, + "num_input_tokens_seen": 231776, + "step": 870 + }, + { + "epoch": 0.22935779816513763, + "grad_norm": 5.481513977050781, + "learning_rate": 1.145478374836173e-05, + "loss": 0.7968, + "num_input_tokens_seen": 233104, + "step": 875 + }, + { + "epoch": 0.2306684141546527, + "grad_norm": 13.893253326416016, + "learning_rate": 1.1520314547837483e-05, + "loss": 0.6053, + "num_input_tokens_seen": 234128, + "step": 880 + }, + { + "epoch": 0.23197903014416776, + "grad_norm": 5.439612865447998, + "learning_rate": 1.1585845347313237e-05, + "loss": 0.6386, + "num_input_tokens_seen": 235312, + "step": 885 + }, + { + "epoch": 0.23328964613368283, + "grad_norm": 9.216582298278809, + "learning_rate": 1.1651376146788991e-05, + "loss": 0.5435, + "num_input_tokens_seen": 236896, + "step": 890 + }, + { + "epoch": 0.2346002621231979, + "grad_norm": 2.4130938053131104, + "learning_rate": 1.1716906946264745e-05, + "loss": 0.621, + "num_input_tokens_seen": 237936, + "step": 895 + }, + { + "epoch": 0.23591087811271297, + "grad_norm": 3.2513976097106934, + "learning_rate": 1.1782437745740499e-05, + "loss": 0.6815, + "num_input_tokens_seen": 239552, + "step": 900 + }, + { + "epoch": 0.23722149410222804, + "grad_norm": 6.407278060913086, + "learning_rate": 1.1847968545216253e-05, + "loss": 0.5094, + "num_input_tokens_seen": 240832, + "step": 905 + }, + { + "epoch": 0.23853211009174313, + "grad_norm": 4.541834831237793, + "learning_rate": 1.1913499344692005e-05, + "loss": 0.618, + "num_input_tokens_seen": 242192, + "step": 910 + }, + { + "epoch": 0.2398427260812582, + "grad_norm": 7.36311674118042, + "learning_rate": 1.197903014416776e-05, + "loss": 0.6321, + "num_input_tokens_seen": 243456, + "step": 915 + }, + { + "epoch": 0.24115334207077327, + "grad_norm": 4.196935653686523, + "learning_rate": 1.2044560943643513e-05, + "loss": 0.7367, + "num_input_tokens_seen": 245424, + "step": 920 + }, + { + "epoch": 0.24246395806028834, + "grad_norm": 2.8827927112579346, + "learning_rate": 1.2110091743119267e-05, + "loss": 0.5099, + "num_input_tokens_seen": 246752, + "step": 925 + }, + { + "epoch": 0.2437745740498034, + "grad_norm": 2.0395805835723877, + "learning_rate": 1.2175622542595021e-05, + "loss": 0.5029, + "num_input_tokens_seen": 248256, + "step": 930 + }, + { + "epoch": 0.24508519003931847, + "grad_norm": 6.901983261108398, + "learning_rate": 1.2241153342070775e-05, + "loss": 0.501, + "num_input_tokens_seen": 249648, + "step": 935 + }, + { + "epoch": 0.24639580602883354, + "grad_norm": 13.09288215637207, + "learning_rate": 1.2306684141546527e-05, + "loss": 0.652, + "num_input_tokens_seen": 251600, + "step": 940 + }, + { + "epoch": 0.24770642201834864, + "grad_norm": 2.529425621032715, + "learning_rate": 1.2372214941022281e-05, + "loss": 0.5978, + "num_input_tokens_seen": 252544, + "step": 945 + }, + { + "epoch": 0.2490170380078637, + "grad_norm": 5.80247163772583, + "learning_rate": 1.2437745740498035e-05, + "loss": 1.0158, + "num_input_tokens_seen": 253776, + "step": 950 + }, + { + "epoch": 0.2503276539973788, + "grad_norm": 9.45703125, + "learning_rate": 1.2503276539973787e-05, + "loss": 0.4462, + "num_input_tokens_seen": 255232, + "step": 955 + }, + { + "epoch": 0.2516382699868938, + "grad_norm": 5.641648292541504, + "learning_rate": 1.2568807339449543e-05, + "loss": 0.6149, + "num_input_tokens_seen": 256640, + "step": 960 + }, + { + "epoch": 0.2529488859764089, + "grad_norm": 12.18066692352295, + "learning_rate": 1.2634338138925295e-05, + "loss": 0.5854, + "num_input_tokens_seen": 257760, + "step": 965 + }, + { + "epoch": 0.254259501965924, + "grad_norm": 5.092421054840088, + "learning_rate": 1.2699868938401049e-05, + "loss": 0.4989, + "num_input_tokens_seen": 258960, + "step": 970 + }, + { + "epoch": 0.25557011795543905, + "grad_norm": 12.951010704040527, + "learning_rate": 1.2765399737876801e-05, + "loss": 0.4653, + "num_input_tokens_seen": 260528, + "step": 975 + }, + { + "epoch": 0.25688073394495414, + "grad_norm": 2.9334073066711426, + "learning_rate": 1.2830930537352557e-05, + "loss": 0.5024, + "num_input_tokens_seen": 261744, + "step": 980 + }, + { + "epoch": 0.2581913499344692, + "grad_norm": 3.943568706512451, + "learning_rate": 1.289646133682831e-05, + "loss": 0.6827, + "num_input_tokens_seen": 262928, + "step": 985 + }, + { + "epoch": 0.2595019659239843, + "grad_norm": 1.9689085483551025, + "learning_rate": 1.2961992136304063e-05, + "loss": 0.5349, + "num_input_tokens_seen": 264384, + "step": 990 + }, + { + "epoch": 0.2608125819134993, + "grad_norm": 46.51886749267578, + "learning_rate": 1.3027522935779818e-05, + "loss": 0.9413, + "num_input_tokens_seen": 265392, + "step": 995 + }, + { + "epoch": 0.2621231979030144, + "grad_norm": 13.375801086425781, + "learning_rate": 1.309305373525557e-05, + "loss": 0.8985, + "num_input_tokens_seen": 266608, + "step": 1000 + }, + { + "epoch": 0.2634338138925295, + "grad_norm": 4.388759613037109, + "learning_rate": 1.3158584534731325e-05, + "loss": 0.5641, + "num_input_tokens_seen": 267984, + "step": 1005 + }, + { + "epoch": 0.26474442988204455, + "grad_norm": 7.730613708496094, + "learning_rate": 1.3224115334207077e-05, + "loss": 0.7562, + "num_input_tokens_seen": 268944, + "step": 1010 + }, + { + "epoch": 0.26605504587155965, + "grad_norm": 3.07076358795166, + "learning_rate": 1.328964613368283e-05, + "loss": 0.5354, + "num_input_tokens_seen": 270672, + "step": 1015 + }, + { + "epoch": 0.2673656618610747, + "grad_norm": 5.243984222412109, + "learning_rate": 1.3355176933158586e-05, + "loss": 0.483, + "num_input_tokens_seen": 272208, + "step": 1020 + }, + { + "epoch": 0.2686762778505898, + "grad_norm": 3.9766640663146973, + "learning_rate": 1.3420707732634339e-05, + "loss": 0.6171, + "num_input_tokens_seen": 274720, + "step": 1025 + }, + { + "epoch": 0.2699868938401048, + "grad_norm": 153.75033569335938, + "learning_rate": 1.3486238532110092e-05, + "loss": 2.272, + "num_input_tokens_seen": 275664, + "step": 1030 + }, + { + "epoch": 0.2712975098296199, + "grad_norm": 5.889108180999756, + "learning_rate": 1.3551769331585845e-05, + "loss": 0.7778, + "num_input_tokens_seen": 277568, + "step": 1035 + }, + { + "epoch": 0.272608125819135, + "grad_norm": 2.787822961807251, + "learning_rate": 1.36173001310616e-05, + "loss": 0.493, + "num_input_tokens_seen": 278848, + "step": 1040 + }, + { + "epoch": 0.27391874180865006, + "grad_norm": 1.9986153841018677, + "learning_rate": 1.3682830930537352e-05, + "loss": 0.4935, + "num_input_tokens_seen": 279968, + "step": 1045 + }, + { + "epoch": 0.27522935779816515, + "grad_norm": 4.190087795257568, + "learning_rate": 1.3748361730013106e-05, + "loss": 0.8195, + "num_input_tokens_seen": 281184, + "step": 1050 + }, + { + "epoch": 0.2765399737876802, + "grad_norm": 2.8889832496643066, + "learning_rate": 1.3813892529488862e-05, + "loss": 0.4179, + "num_input_tokens_seen": 282576, + "step": 1055 + }, + { + "epoch": 0.2778505897771953, + "grad_norm": 2.48128342628479, + "learning_rate": 1.3879423328964614e-05, + "loss": 0.4908, + "num_input_tokens_seen": 284064, + "step": 1060 + }, + { + "epoch": 0.27916120576671033, + "grad_norm": 17.967483520507812, + "learning_rate": 1.3944954128440368e-05, + "loss": 0.5646, + "num_input_tokens_seen": 285200, + "step": 1065 + }, + { + "epoch": 0.2804718217562254, + "grad_norm": 5.220232963562012, + "learning_rate": 1.401048492791612e-05, + "loss": 0.5693, + "num_input_tokens_seen": 286432, + "step": 1070 + }, + { + "epoch": 0.2817824377457405, + "grad_norm": 3.1587166786193848, + "learning_rate": 1.4076015727391876e-05, + "loss": 0.5216, + "num_input_tokens_seen": 287888, + "step": 1075 + }, + { + "epoch": 0.28309305373525556, + "grad_norm": 7.073264122009277, + "learning_rate": 1.4141546526867626e-05, + "loss": 0.6007, + "num_input_tokens_seen": 289120, + "step": 1080 + }, + { + "epoch": 0.28440366972477066, + "grad_norm": 4.988691806793213, + "learning_rate": 1.4207077326343382e-05, + "loss": 0.7544, + "num_input_tokens_seen": 290032, + "step": 1085 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 4.278416156768799, + "learning_rate": 1.4272608125819138e-05, + "loss": 0.4716, + "num_input_tokens_seen": 291328, + "step": 1090 + }, + { + "epoch": 0.2870249017038008, + "grad_norm": 2.0615153312683105, + "learning_rate": 1.4338138925294888e-05, + "loss": 0.3886, + "num_input_tokens_seen": 293632, + "step": 1095 + }, + { + "epoch": 0.28833551769331583, + "grad_norm": 3.0684690475463867, + "learning_rate": 1.4403669724770644e-05, + "loss": 0.734, + "num_input_tokens_seen": 294880, + "step": 1100 + }, + { + "epoch": 0.28964613368283093, + "grad_norm": 3.7555453777313232, + "learning_rate": 1.4469200524246396e-05, + "loss": 0.5104, + "num_input_tokens_seen": 296208, + "step": 1105 + }, + { + "epoch": 0.290956749672346, + "grad_norm": 1.8215569257736206, + "learning_rate": 1.453473132372215e-05, + "loss": 0.5594, + "num_input_tokens_seen": 297760, + "step": 1110 + }, + { + "epoch": 0.29226736566186107, + "grad_norm": 5.37257194519043, + "learning_rate": 1.4600262123197902e-05, + "loss": 0.5377, + "num_input_tokens_seen": 298832, + "step": 1115 + }, + { + "epoch": 0.29357798165137616, + "grad_norm": 2.6148552894592285, + "learning_rate": 1.4665792922673658e-05, + "loss": 0.559, + "num_input_tokens_seen": 300112, + "step": 1120 + }, + { + "epoch": 0.2948885976408912, + "grad_norm": 3.59912371635437, + "learning_rate": 1.4731323722149412e-05, + "loss": 0.634, + "num_input_tokens_seen": 301248, + "step": 1125 + }, + { + "epoch": 0.2961992136304063, + "grad_norm": 4.620095252990723, + "learning_rate": 1.4796854521625164e-05, + "loss": 0.5054, + "num_input_tokens_seen": 302656, + "step": 1130 + }, + { + "epoch": 0.29750982961992134, + "grad_norm": 2.7268104553222656, + "learning_rate": 1.486238532110092e-05, + "loss": 0.5848, + "num_input_tokens_seen": 304000, + "step": 1135 + }, + { + "epoch": 0.29882044560943644, + "grad_norm": 3.36013126373291, + "learning_rate": 1.4927916120576672e-05, + "loss": 0.6479, + "num_input_tokens_seen": 305472, + "step": 1140 + }, + { + "epoch": 0.30013106159895153, + "grad_norm": 5.485869884490967, + "learning_rate": 1.4993446920052426e-05, + "loss": 0.345, + "num_input_tokens_seen": 306448, + "step": 1145 + }, + { + "epoch": 0.30144167758846657, + "grad_norm": 18.646595001220703, + "learning_rate": 1.5058977719528178e-05, + "loss": 0.8363, + "num_input_tokens_seen": 307632, + "step": 1150 + }, + { + "epoch": 0.30275229357798167, + "grad_norm": 2.3952248096466064, + "learning_rate": 1.5124508519003932e-05, + "loss": 0.3954, + "num_input_tokens_seen": 308784, + "step": 1155 + }, + { + "epoch": 0.3040629095674967, + "grad_norm": 10.275803565979004, + "learning_rate": 1.5190039318479687e-05, + "loss": 0.5174, + "num_input_tokens_seen": 310096, + "step": 1160 + }, + { + "epoch": 0.3053735255570118, + "grad_norm": 3.9282500743865967, + "learning_rate": 1.525557011795544e-05, + "loss": 1.2299, + "num_input_tokens_seen": 311472, + "step": 1165 + }, + { + "epoch": 0.30668414154652685, + "grad_norm": 1.5429003238677979, + "learning_rate": 1.5321100917431195e-05, + "loss": 0.6401, + "num_input_tokens_seen": 312704, + "step": 1170 + }, + { + "epoch": 0.30799475753604194, + "grad_norm": 4.482446193695068, + "learning_rate": 1.5386631716906946e-05, + "loss": 0.4761, + "num_input_tokens_seen": 313888, + "step": 1175 + }, + { + "epoch": 0.30930537352555704, + "grad_norm": 2.115330457687378, + "learning_rate": 1.54521625163827e-05, + "loss": 0.5755, + "num_input_tokens_seen": 315040, + "step": 1180 + }, + { + "epoch": 0.3106159895150721, + "grad_norm": 3.3864316940307617, + "learning_rate": 1.5517693315858454e-05, + "loss": 0.9387, + "num_input_tokens_seen": 316352, + "step": 1185 + }, + { + "epoch": 0.3119266055045872, + "grad_norm": 3.83917236328125, + "learning_rate": 1.5583224115334208e-05, + "loss": 0.6202, + "num_input_tokens_seen": 317616, + "step": 1190 + }, + { + "epoch": 0.3132372214941022, + "grad_norm": 3.6829774379730225, + "learning_rate": 1.564875491480996e-05, + "loss": 0.6555, + "num_input_tokens_seen": 319184, + "step": 1195 + }, + { + "epoch": 0.3145478374836173, + "grad_norm": 3.715125322341919, + "learning_rate": 1.5714285714285715e-05, + "loss": 0.5981, + "num_input_tokens_seen": 320752, + "step": 1200 + }, + { + "epoch": 0.31585845347313235, + "grad_norm": 5.258845806121826, + "learning_rate": 1.577981651376147e-05, + "loss": 0.4747, + "num_input_tokens_seen": 322032, + "step": 1205 + }, + { + "epoch": 0.31716906946264745, + "grad_norm": 2.707841634750366, + "learning_rate": 1.5845347313237223e-05, + "loss": 0.5116, + "num_input_tokens_seen": 323360, + "step": 1210 + }, + { + "epoch": 0.31847968545216254, + "grad_norm": 10.32327938079834, + "learning_rate": 1.5910878112712977e-05, + "loss": 0.473, + "num_input_tokens_seen": 324528, + "step": 1215 + }, + { + "epoch": 0.3197903014416776, + "grad_norm": 7.717580795288086, + "learning_rate": 1.5976408912188728e-05, + "loss": 0.5086, + "num_input_tokens_seen": 326928, + "step": 1220 + }, + { + "epoch": 0.3211009174311927, + "grad_norm": 8.760111808776855, + "learning_rate": 1.604193971166448e-05, + "loss": 0.4242, + "num_input_tokens_seen": 328432, + "step": 1225 + }, + { + "epoch": 0.3224115334207077, + "grad_norm": 10.228850364685059, + "learning_rate": 1.610747051114024e-05, + "loss": 0.5819, + "num_input_tokens_seen": 329440, + "step": 1230 + }, + { + "epoch": 0.3237221494102228, + "grad_norm": 3.0171453952789307, + "learning_rate": 1.617300131061599e-05, + "loss": 0.3952, + "num_input_tokens_seen": 330912, + "step": 1235 + }, + { + "epoch": 0.32503276539973786, + "grad_norm": 2.4059982299804688, + "learning_rate": 1.6238532110091743e-05, + "loss": 0.5953, + "num_input_tokens_seen": 332288, + "step": 1240 + }, + { + "epoch": 0.32634338138925295, + "grad_norm": 3.0065600872039795, + "learning_rate": 1.6304062909567497e-05, + "loss": 0.5915, + "num_input_tokens_seen": 333584, + "step": 1245 + }, + { + "epoch": 0.32765399737876805, + "grad_norm": 2.064418315887451, + "learning_rate": 1.636959370904325e-05, + "loss": 0.5593, + "num_input_tokens_seen": 335392, + "step": 1250 + }, + { + "epoch": 0.3289646133682831, + "grad_norm": 3.591035842895508, + "learning_rate": 1.6435124508519005e-05, + "loss": 0.6625, + "num_input_tokens_seen": 336560, + "step": 1255 + }, + { + "epoch": 0.3302752293577982, + "grad_norm": 2.939518451690674, + "learning_rate": 1.650065530799476e-05, + "loss": 0.634, + "num_input_tokens_seen": 338192, + "step": 1260 + }, + { + "epoch": 0.3315858453473132, + "grad_norm": 4.092834949493408, + "learning_rate": 1.6566186107470513e-05, + "loss": 0.6497, + "num_input_tokens_seen": 339568, + "step": 1265 + }, + { + "epoch": 0.3328964613368283, + "grad_norm": 8.921924591064453, + "learning_rate": 1.6631716906946267e-05, + "loss": 0.4776, + "num_input_tokens_seen": 340736, + "step": 1270 + }, + { + "epoch": 0.33420707732634336, + "grad_norm": 5.908846855163574, + "learning_rate": 1.669724770642202e-05, + "loss": 1.1958, + "num_input_tokens_seen": 341680, + "step": 1275 + }, + { + "epoch": 0.33551769331585846, + "grad_norm": 2.470532178878784, + "learning_rate": 1.676277850589777e-05, + "loss": 0.6276, + "num_input_tokens_seen": 342800, + "step": 1280 + }, + { + "epoch": 0.33682830930537355, + "grad_norm": 2.3337106704711914, + "learning_rate": 1.682830930537353e-05, + "loss": 0.6474, + "num_input_tokens_seen": 344000, + "step": 1285 + }, + { + "epoch": 0.3381389252948886, + "grad_norm": 3.9827418327331543, + "learning_rate": 1.689384010484928e-05, + "loss": 0.5122, + "num_input_tokens_seen": 345968, + "step": 1290 + }, + { + "epoch": 0.3394495412844037, + "grad_norm": 5.9223313331604, + "learning_rate": 1.6959370904325033e-05, + "loss": 0.4873, + "num_input_tokens_seen": 347264, + "step": 1295 + }, + { + "epoch": 0.34076015727391873, + "grad_norm": 31.87460708618164, + "learning_rate": 1.702490170380079e-05, + "loss": 0.5323, + "num_input_tokens_seen": 348720, + "step": 1300 + }, + { + "epoch": 0.3420707732634338, + "grad_norm": 3.0745654106140137, + "learning_rate": 1.709043250327654e-05, + "loss": 0.43, + "num_input_tokens_seen": 350032, + "step": 1305 + }, + { + "epoch": 0.34338138925294887, + "grad_norm": 2.9798035621643066, + "learning_rate": 1.7155963302752295e-05, + "loss": 0.5165, + "num_input_tokens_seen": 351408, + "step": 1310 + }, + { + "epoch": 0.34469200524246396, + "grad_norm": 1.6038254499435425, + "learning_rate": 1.722149410222805e-05, + "loss": 0.4325, + "num_input_tokens_seen": 352560, + "step": 1315 + }, + { + "epoch": 0.34600262123197906, + "grad_norm": 4.227873802185059, + "learning_rate": 1.7287024901703802e-05, + "loss": 0.5998, + "num_input_tokens_seen": 353824, + "step": 1320 + }, + { + "epoch": 0.3473132372214941, + "grad_norm": 3.0901293754577637, + "learning_rate": 1.7352555701179553e-05, + "loss": 0.4273, + "num_input_tokens_seen": 355456, + "step": 1325 + }, + { + "epoch": 0.3486238532110092, + "grad_norm": 1.8432401418685913, + "learning_rate": 1.741808650065531e-05, + "loss": 0.4973, + "num_input_tokens_seen": 356528, + "step": 1330 + }, + { + "epoch": 0.34993446920052423, + "grad_norm": 2.7687149047851562, + "learning_rate": 1.7483617300131064e-05, + "loss": 0.3728, + "num_input_tokens_seen": 357664, + "step": 1335 + }, + { + "epoch": 0.35124508519003933, + "grad_norm": 1.8252090215682983, + "learning_rate": 1.7549148099606815e-05, + "loss": 0.6563, + "num_input_tokens_seen": 359136, + "step": 1340 + }, + { + "epoch": 0.35255570117955437, + "grad_norm": 2.2599785327911377, + "learning_rate": 1.7614678899082572e-05, + "loss": 0.3813, + "num_input_tokens_seen": 360560, + "step": 1345 + }, + { + "epoch": 0.35386631716906947, + "grad_norm": 2.1658895015716553, + "learning_rate": 1.7680209698558323e-05, + "loss": 0.6403, + "num_input_tokens_seen": 361952, + "step": 1350 + }, + { + "epoch": 0.35517693315858456, + "grad_norm": 6.3700480461120605, + "learning_rate": 1.7745740498034076e-05, + "loss": 0.5332, + "num_input_tokens_seen": 363360, + "step": 1355 + }, + { + "epoch": 0.3564875491480996, + "grad_norm": 2.268155574798584, + "learning_rate": 1.781127129750983e-05, + "loss": 0.6022, + "num_input_tokens_seen": 364480, + "step": 1360 + }, + { + "epoch": 0.3577981651376147, + "grad_norm": 15.67795467376709, + "learning_rate": 1.7876802096985584e-05, + "loss": 0.7471, + "num_input_tokens_seen": 365584, + "step": 1365 + }, + { + "epoch": 0.35910878112712974, + "grad_norm": 2.787968158721924, + "learning_rate": 1.7942332896461335e-05, + "loss": 0.7135, + "num_input_tokens_seen": 366704, + "step": 1370 + }, + { + "epoch": 0.36041939711664484, + "grad_norm": 2.7672431468963623, + "learning_rate": 1.8007863695937092e-05, + "loss": 0.7771, + "num_input_tokens_seen": 368000, + "step": 1375 + }, + { + "epoch": 0.3617300131061599, + "grad_norm": 4.215776443481445, + "learning_rate": 1.8073394495412846e-05, + "loss": 0.4134, + "num_input_tokens_seen": 369360, + "step": 1380 + }, + { + "epoch": 0.36304062909567497, + "grad_norm": 1.7084484100341797, + "learning_rate": 1.8138925294888597e-05, + "loss": 0.53, + "num_input_tokens_seen": 370864, + "step": 1385 + }, + { + "epoch": 0.36435124508519, + "grad_norm": 3.618713140487671, + "learning_rate": 1.8204456094364354e-05, + "loss": 0.3567, + "num_input_tokens_seen": 372000, + "step": 1390 + }, + { + "epoch": 0.3656618610747051, + "grad_norm": 78.53734588623047, + "learning_rate": 1.8269986893840104e-05, + "loss": 0.9822, + "num_input_tokens_seen": 373168, + "step": 1395 + }, + { + "epoch": 0.3669724770642202, + "grad_norm": 5.159477233886719, + "learning_rate": 1.833551769331586e-05, + "loss": 0.64, + "num_input_tokens_seen": 374304, + "step": 1400 + }, + { + "epoch": 0.36828309305373524, + "grad_norm": 13.948582649230957, + "learning_rate": 1.8401048492791612e-05, + "loss": 0.3163, + "num_input_tokens_seen": 375888, + "step": 1405 + }, + { + "epoch": 0.36959370904325034, + "grad_norm": 17.960376739501953, + "learning_rate": 1.8466579292267366e-05, + "loss": 0.5838, + "num_input_tokens_seen": 377296, + "step": 1410 + }, + { + "epoch": 0.3709043250327654, + "grad_norm": 4.355318069458008, + "learning_rate": 1.853211009174312e-05, + "loss": 0.4152, + "num_input_tokens_seen": 378608, + "step": 1415 + }, + { + "epoch": 0.3722149410222805, + "grad_norm": 1.7192683219909668, + "learning_rate": 1.8597640891218874e-05, + "loss": 0.4893, + "num_input_tokens_seen": 380192, + "step": 1420 + }, + { + "epoch": 0.3735255570117955, + "grad_norm": 2.8599560260772705, + "learning_rate": 1.8663171690694628e-05, + "loss": 0.5901, + "num_input_tokens_seen": 381616, + "step": 1425 + }, + { + "epoch": 0.3748361730013106, + "grad_norm": 3.0564002990722656, + "learning_rate": 1.872870249017038e-05, + "loss": 0.505, + "num_input_tokens_seen": 382672, + "step": 1430 + }, + { + "epoch": 0.3761467889908257, + "grad_norm": 4.439542770385742, + "learning_rate": 1.8794233289646136e-05, + "loss": 0.5805, + "num_input_tokens_seen": 384176, + "step": 1435 + }, + { + "epoch": 0.37745740498034075, + "grad_norm": 2.168962240219116, + "learning_rate": 1.8859764089121886e-05, + "loss": 0.7246, + "num_input_tokens_seen": 385584, + "step": 1440 + }, + { + "epoch": 0.37876802096985585, + "grad_norm": 2.6855225563049316, + "learning_rate": 1.892529488859764e-05, + "loss": 0.4377, + "num_input_tokens_seen": 386832, + "step": 1445 + }, + { + "epoch": 0.3800786369593709, + "grad_norm": 1.9271728992462158, + "learning_rate": 1.8990825688073397e-05, + "loss": 0.4578, + "num_input_tokens_seen": 388416, + "step": 1450 + }, + { + "epoch": 0.381389252948886, + "grad_norm": 3.65674090385437, + "learning_rate": 1.9056356487549148e-05, + "loss": 0.3723, + "num_input_tokens_seen": 389648, + "step": 1455 + }, + { + "epoch": 0.382699868938401, + "grad_norm": 2.1009905338287354, + "learning_rate": 1.9121887287024902e-05, + "loss": 0.3484, + "num_input_tokens_seen": 390864, + "step": 1460 + }, + { + "epoch": 0.3840104849279161, + "grad_norm": 2.4598424434661865, + "learning_rate": 1.9187418086500656e-05, + "loss": 0.5137, + "num_input_tokens_seen": 392128, + "step": 1465 + }, + { + "epoch": 0.3853211009174312, + "grad_norm": 2.568999767303467, + "learning_rate": 1.925294888597641e-05, + "loss": 0.6429, + "num_input_tokens_seen": 394864, + "step": 1470 + }, + { + "epoch": 0.38663171690694625, + "grad_norm": 8.142760276794434, + "learning_rate": 1.9318479685452164e-05, + "loss": 0.4161, + "num_input_tokens_seen": 395776, + "step": 1475 + }, + { + "epoch": 0.38794233289646135, + "grad_norm": 2.960918426513672, + "learning_rate": 1.9384010484927918e-05, + "loss": 0.7234, + "num_input_tokens_seen": 396880, + "step": 1480 + }, + { + "epoch": 0.3892529488859764, + "grad_norm": 2.9498050212860107, + "learning_rate": 1.944954128440367e-05, + "loss": 0.4897, + "num_input_tokens_seen": 398512, + "step": 1485 + }, + { + "epoch": 0.3905635648754915, + "grad_norm": 6.700377464294434, + "learning_rate": 1.9515072083879425e-05, + "loss": 0.7631, + "num_input_tokens_seen": 399664, + "step": 1490 + }, + { + "epoch": 0.3918741808650065, + "grad_norm": 5.012332439422607, + "learning_rate": 1.958060288335518e-05, + "loss": 0.3469, + "num_input_tokens_seen": 400864, + "step": 1495 + }, + { + "epoch": 0.3931847968545216, + "grad_norm": 1.8109019994735718, + "learning_rate": 1.964613368283093e-05, + "loss": 0.5932, + "num_input_tokens_seen": 402352, + "step": 1500 + }, + { + "epoch": 0.3944954128440367, + "grad_norm": 2.0942559242248535, + "learning_rate": 1.9711664482306684e-05, + "loss": 0.5632, + "num_input_tokens_seen": 403744, + "step": 1505 + }, + { + "epoch": 0.39580602883355176, + "grad_norm": 6.298874855041504, + "learning_rate": 1.9777195281782438e-05, + "loss": 0.5145, + "num_input_tokens_seen": 405104, + "step": 1510 + }, + { + "epoch": 0.39711664482306686, + "grad_norm": 7.41862154006958, + "learning_rate": 1.984272608125819e-05, + "loss": 0.7461, + "num_input_tokens_seen": 406208, + "step": 1515 + }, + { + "epoch": 0.3984272608125819, + "grad_norm": 9.361674308776855, + "learning_rate": 1.9908256880733945e-05, + "loss": 0.5303, + "num_input_tokens_seen": 407120, + "step": 1520 + }, + { + "epoch": 0.399737876802097, + "grad_norm": 2.468761682510376, + "learning_rate": 1.99737876802097e-05, + "loss": 0.5914, + "num_input_tokens_seen": 408080, + "step": 1525 + }, + { + "epoch": 0.40104849279161203, + "grad_norm": 2.047950267791748, + "learning_rate": 2.0039318479685453e-05, + "loss": 0.5925, + "num_input_tokens_seen": 409360, + "step": 1530 + }, + { + "epoch": 0.40235910878112713, + "grad_norm": 2.6260223388671875, + "learning_rate": 2.0104849279161207e-05, + "loss": 0.4984, + "num_input_tokens_seen": 410720, + "step": 1535 + }, + { + "epoch": 0.4036697247706422, + "grad_norm": 2.7161340713500977, + "learning_rate": 2.017038007863696e-05, + "loss": 0.5547, + "num_input_tokens_seen": 411904, + "step": 1540 + }, + { + "epoch": 0.40498034076015726, + "grad_norm": 2.5378544330596924, + "learning_rate": 2.023591087811271e-05, + "loss": 0.574, + "num_input_tokens_seen": 412960, + "step": 1545 + }, + { + "epoch": 0.40629095674967236, + "grad_norm": 2.976605176925659, + "learning_rate": 2.030144167758847e-05, + "loss": 0.5387, + "num_input_tokens_seen": 414784, + "step": 1550 + }, + { + "epoch": 0.4076015727391874, + "grad_norm": 2.0496301651000977, + "learning_rate": 2.0366972477064223e-05, + "loss": 0.5704, + "num_input_tokens_seen": 416208, + "step": 1555 + }, + { + "epoch": 0.4089121887287025, + "grad_norm": 2.687800168991089, + "learning_rate": 2.0432503276539973e-05, + "loss": 0.4686, + "num_input_tokens_seen": 417440, + "step": 1560 + }, + { + "epoch": 0.41022280471821754, + "grad_norm": 1.3792091608047485, + "learning_rate": 2.049803407601573e-05, + "loss": 0.5881, + "num_input_tokens_seen": 419008, + "step": 1565 + }, + { + "epoch": 0.41153342070773263, + "grad_norm": 2.484091281890869, + "learning_rate": 2.056356487549148e-05, + "loss": 0.5021, + "num_input_tokens_seen": 420384, + "step": 1570 + }, + { + "epoch": 0.41284403669724773, + "grad_norm": 5.72295618057251, + "learning_rate": 2.0629095674967235e-05, + "loss": 0.5988, + "num_input_tokens_seen": 421440, + "step": 1575 + }, + { + "epoch": 0.41415465268676277, + "grad_norm": 2.1795554161071777, + "learning_rate": 2.069462647444299e-05, + "loss": 0.5494, + "num_input_tokens_seen": 422608, + "step": 1580 + }, + { + "epoch": 0.41546526867627787, + "grad_norm": 2.938673734664917, + "learning_rate": 2.0760157273918743e-05, + "loss": 0.527, + "num_input_tokens_seen": 423952, + "step": 1585 + }, + { + "epoch": 0.4167758846657929, + "grad_norm": 2.0420374870300293, + "learning_rate": 2.0825688073394497e-05, + "loss": 0.5064, + "num_input_tokens_seen": 425552, + "step": 1590 + }, + { + "epoch": 0.418086500655308, + "grad_norm": 3.705045223236084, + "learning_rate": 2.089121887287025e-05, + "loss": 0.6502, + "num_input_tokens_seen": 426656, + "step": 1595 + }, + { + "epoch": 0.41939711664482304, + "grad_norm": 1.6153628826141357, + "learning_rate": 2.0956749672346005e-05, + "loss": 0.7287, + "num_input_tokens_seen": 427696, + "step": 1600 + }, + { + "epoch": 0.42070773263433814, + "grad_norm": 3.5930697917938232, + "learning_rate": 2.1022280471821755e-05, + "loss": 0.5959, + "num_input_tokens_seen": 428656, + "step": 1605 + }, + { + "epoch": 0.42201834862385323, + "grad_norm": 2.2805685997009277, + "learning_rate": 2.1087811271297513e-05, + "loss": 0.6838, + "num_input_tokens_seen": 430128, + "step": 1610 + }, + { + "epoch": 0.4233289646133683, + "grad_norm": 1.792470097541809, + "learning_rate": 2.1153342070773263e-05, + "loss": 0.3815, + "num_input_tokens_seen": 431792, + "step": 1615 + }, + { + "epoch": 0.42463958060288337, + "grad_norm": 9.89884090423584, + "learning_rate": 2.1218872870249017e-05, + "loss": 0.4181, + "num_input_tokens_seen": 433200, + "step": 1620 + }, + { + "epoch": 0.4259501965923984, + "grad_norm": 2.018691062927246, + "learning_rate": 2.1284403669724774e-05, + "loss": 0.4455, + "num_input_tokens_seen": 434688, + "step": 1625 + }, + { + "epoch": 0.4272608125819135, + "grad_norm": 2.6907851696014404, + "learning_rate": 2.1349934469200525e-05, + "loss": 0.7561, + "num_input_tokens_seen": 435920, + "step": 1630 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 2.13456392288208, + "learning_rate": 2.141546526867628e-05, + "loss": 0.5686, + "num_input_tokens_seen": 438000, + "step": 1635 + }, + { + "epoch": 0.42988204456094364, + "grad_norm": 2.8843135833740234, + "learning_rate": 2.1480996068152033e-05, + "loss": 0.5332, + "num_input_tokens_seen": 439104, + "step": 1640 + }, + { + "epoch": 0.43119266055045874, + "grad_norm": 2.2328808307647705, + "learning_rate": 2.1546526867627787e-05, + "loss": 0.5709, + "num_input_tokens_seen": 440400, + "step": 1645 + }, + { + "epoch": 0.4325032765399738, + "grad_norm": 3.585833787918091, + "learning_rate": 2.1612057667103537e-05, + "loss": 0.5826, + "num_input_tokens_seen": 442592, + "step": 1650 + }, + { + "epoch": 0.4338138925294889, + "grad_norm": 3.6883668899536133, + "learning_rate": 2.1677588466579294e-05, + "loss": 0.6349, + "num_input_tokens_seen": 444016, + "step": 1655 + }, + { + "epoch": 0.4351245085190039, + "grad_norm": 3.7169089317321777, + "learning_rate": 2.1743119266055048e-05, + "loss": 0.5075, + "num_input_tokens_seen": 445424, + "step": 1660 + }, + { + "epoch": 0.436435124508519, + "grad_norm": 1.3879845142364502, + "learning_rate": 2.18086500655308e-05, + "loss": 0.4579, + "num_input_tokens_seen": 446768, + "step": 1665 + }, + { + "epoch": 0.43774574049803405, + "grad_norm": 2.0107383728027344, + "learning_rate": 2.1874180865006556e-05, + "loss": 0.7532, + "num_input_tokens_seen": 448288, + "step": 1670 + }, + { + "epoch": 0.43905635648754915, + "grad_norm": 2.6797235012054443, + "learning_rate": 2.1939711664482307e-05, + "loss": 0.7297, + "num_input_tokens_seen": 449440, + "step": 1675 + }, + { + "epoch": 0.44036697247706424, + "grad_norm": 4.476974964141846, + "learning_rate": 2.200524246395806e-05, + "loss": 0.4648, + "num_input_tokens_seen": 450832, + "step": 1680 + }, + { + "epoch": 0.4416775884665793, + "grad_norm": 2.445460081100464, + "learning_rate": 2.2070773263433814e-05, + "loss": 0.6376, + "num_input_tokens_seen": 451984, + "step": 1685 + }, + { + "epoch": 0.4429882044560944, + "grad_norm": 2.4229190349578857, + "learning_rate": 2.213630406290957e-05, + "loss": 0.5312, + "num_input_tokens_seen": 453376, + "step": 1690 + }, + { + "epoch": 0.4442988204456094, + "grad_norm": 2.843794822692871, + "learning_rate": 2.2201834862385322e-05, + "loss": 0.424, + "num_input_tokens_seen": 454672, + "step": 1695 + }, + { + "epoch": 0.4456094364351245, + "grad_norm": 1.4180983304977417, + "learning_rate": 2.2267365661861076e-05, + "loss": 0.4403, + "num_input_tokens_seen": 456240, + "step": 1700 + }, + { + "epoch": 0.44692005242463956, + "grad_norm": 14.49243450164795, + "learning_rate": 2.233289646133683e-05, + "loss": 0.4756, + "num_input_tokens_seen": 457632, + "step": 1705 + }, + { + "epoch": 0.44823066841415465, + "grad_norm": 6.1120429039001465, + "learning_rate": 2.239842726081258e-05, + "loss": 0.3792, + "num_input_tokens_seen": 459200, + "step": 1710 + }, + { + "epoch": 0.44954128440366975, + "grad_norm": 2.4467358589172363, + "learning_rate": 2.2463958060288338e-05, + "loss": 0.5779, + "num_input_tokens_seen": 460384, + "step": 1715 + }, + { + "epoch": 0.4508519003931848, + "grad_norm": 1.956210970878601, + "learning_rate": 2.252948885976409e-05, + "loss": 0.5067, + "num_input_tokens_seen": 461712, + "step": 1720 + }, + { + "epoch": 0.4521625163826999, + "grad_norm": 2.5175676345825195, + "learning_rate": 2.2595019659239842e-05, + "loss": 0.3813, + "num_input_tokens_seen": 462832, + "step": 1725 + }, + { + "epoch": 0.4534731323722149, + "grad_norm": 1.5763418674468994, + "learning_rate": 2.26605504587156e-05, + "loss": 0.4115, + "num_input_tokens_seen": 464224, + "step": 1730 + }, + { + "epoch": 0.45478374836173, + "grad_norm": 1.0184913873672485, + "learning_rate": 2.272608125819135e-05, + "loss": 0.4886, + "num_input_tokens_seen": 465696, + "step": 1735 + }, + { + "epoch": 0.45609436435124506, + "grad_norm": 2.530060052871704, + "learning_rate": 2.2791612057667104e-05, + "loss": 0.542, + "num_input_tokens_seen": 466944, + "step": 1740 + }, + { + "epoch": 0.45740498034076016, + "grad_norm": 1.6662249565124512, + "learning_rate": 2.2857142857142858e-05, + "loss": 0.5699, + "num_input_tokens_seen": 468256, + "step": 1745 + }, + { + "epoch": 0.45871559633027525, + "grad_norm": 3.4794814586639404, + "learning_rate": 2.2922673656618612e-05, + "loss": 0.5485, + "num_input_tokens_seen": 469376, + "step": 1750 + }, + { + "epoch": 0.4600262123197903, + "grad_norm": 2.834761142730713, + "learning_rate": 2.2988204456094366e-05, + "loss": 0.6248, + "num_input_tokens_seen": 470336, + "step": 1755 + }, + { + "epoch": 0.4613368283093054, + "grad_norm": 1.9460679292678833, + "learning_rate": 2.305373525557012e-05, + "loss": 0.6179, + "num_input_tokens_seen": 471360, + "step": 1760 + }, + { + "epoch": 0.46264744429882043, + "grad_norm": 3.1786699295043945, + "learning_rate": 2.3119266055045874e-05, + "loss": 0.6513, + "num_input_tokens_seen": 472576, + "step": 1765 + }, + { + "epoch": 0.4639580602883355, + "grad_norm": 2.8908395767211914, + "learning_rate": 2.3184796854521628e-05, + "loss": 0.6138, + "num_input_tokens_seen": 474224, + "step": 1770 + }, + { + "epoch": 0.46526867627785057, + "grad_norm": 2.5875468254089355, + "learning_rate": 2.325032765399738e-05, + "loss": 0.5069, + "num_input_tokens_seen": 475376, + "step": 1775 + }, + { + "epoch": 0.46657929226736566, + "grad_norm": 2.733396053314209, + "learning_rate": 2.3315858453473132e-05, + "loss": 0.681, + "num_input_tokens_seen": 476592, + "step": 1780 + }, + { + "epoch": 0.46788990825688076, + "grad_norm": 7.877594470977783, + "learning_rate": 2.3381389252948886e-05, + "loss": 0.6823, + "num_input_tokens_seen": 477504, + "step": 1785 + }, + { + "epoch": 0.4692005242463958, + "grad_norm": 2.960611581802368, + "learning_rate": 2.344692005242464e-05, + "loss": 0.5181, + "num_input_tokens_seen": 478560, + "step": 1790 + }, + { + "epoch": 0.4705111402359109, + "grad_norm": 1.78667414188385, + "learning_rate": 2.3512450851900394e-05, + "loss": 0.4554, + "num_input_tokens_seen": 479920, + "step": 1795 + }, + { + "epoch": 0.47182175622542594, + "grad_norm": 1.2522228956222534, + "learning_rate": 2.3577981651376148e-05, + "loss": 0.3995, + "num_input_tokens_seen": 481264, + "step": 1800 + }, + { + "epoch": 0.47313237221494103, + "grad_norm": 59.30772399902344, + "learning_rate": 2.36435124508519e-05, + "loss": 0.9633, + "num_input_tokens_seen": 482608, + "step": 1805 + }, + { + "epoch": 0.4744429882044561, + "grad_norm": 3.732757329940796, + "learning_rate": 2.3709043250327656e-05, + "loss": 0.4989, + "num_input_tokens_seen": 484096, + "step": 1810 + }, + { + "epoch": 0.47575360419397117, + "grad_norm": 4.890771389007568, + "learning_rate": 2.377457404980341e-05, + "loss": 0.5268, + "num_input_tokens_seen": 485120, + "step": 1815 + }, + { + "epoch": 0.47706422018348627, + "grad_norm": 6.351772308349609, + "learning_rate": 2.3840104849279163e-05, + "loss": 0.3852, + "num_input_tokens_seen": 486080, + "step": 1820 + }, + { + "epoch": 0.4783748361730013, + "grad_norm": 3.46543025970459, + "learning_rate": 2.3905635648754914e-05, + "loss": 0.5655, + "num_input_tokens_seen": 487344, + "step": 1825 + }, + { + "epoch": 0.4796854521625164, + "grad_norm": 8.90031909942627, + "learning_rate": 2.397116644823067e-05, + "loss": 0.4897, + "num_input_tokens_seen": 488608, + "step": 1830 + }, + { + "epoch": 0.48099606815203144, + "grad_norm": 1.85548996925354, + "learning_rate": 2.4036697247706425e-05, + "loss": 0.5226, + "num_input_tokens_seen": 489664, + "step": 1835 + }, + { + "epoch": 0.48230668414154654, + "grad_norm": 2.320826768875122, + "learning_rate": 2.4102228047182176e-05, + "loss": 0.4796, + "num_input_tokens_seen": 491120, + "step": 1840 + }, + { + "epoch": 0.4836173001310616, + "grad_norm": 2.7335915565490723, + "learning_rate": 2.4167758846657933e-05, + "loss": 0.4987, + "num_input_tokens_seen": 492464, + "step": 1845 + }, + { + "epoch": 0.4849279161205767, + "grad_norm": 2.226642370223999, + "learning_rate": 2.4233289646133683e-05, + "loss": 0.5534, + "num_input_tokens_seen": 493712, + "step": 1850 + }, + { + "epoch": 0.48623853211009177, + "grad_norm": 2.224778890609741, + "learning_rate": 2.4298820445609437e-05, + "loss": 0.4145, + "num_input_tokens_seen": 494816, + "step": 1855 + }, + { + "epoch": 0.4875491480996068, + "grad_norm": 1.7492518424987793, + "learning_rate": 2.436435124508519e-05, + "loss": 0.4663, + "num_input_tokens_seen": 496256, + "step": 1860 + }, + { + "epoch": 0.4888597640891219, + "grad_norm": 2.6400959491729736, + "learning_rate": 2.4429882044560945e-05, + "loss": 0.5588, + "num_input_tokens_seen": 497280, + "step": 1865 + }, + { + "epoch": 0.49017038007863695, + "grad_norm": 0.9141639471054077, + "learning_rate": 2.44954128440367e-05, + "loss": 0.6248, + "num_input_tokens_seen": 498144, + "step": 1870 + }, + { + "epoch": 0.49148099606815204, + "grad_norm": 3.1002554893493652, + "learning_rate": 2.4560943643512453e-05, + "loss": 0.5519, + "num_input_tokens_seen": 499648, + "step": 1875 + }, + { + "epoch": 0.4927916120576671, + "grad_norm": 1.4401549100875854, + "learning_rate": 2.4626474442988207e-05, + "loss": 0.4062, + "num_input_tokens_seen": 501488, + "step": 1880 + }, + { + "epoch": 0.4941022280471822, + "grad_norm": 1.7362759113311768, + "learning_rate": 2.4692005242463957e-05, + "loss": 0.6932, + "num_input_tokens_seen": 503456, + "step": 1885 + }, + { + "epoch": 0.4954128440366973, + "grad_norm": 2.438713788986206, + "learning_rate": 2.4757536041939715e-05, + "loss": 0.3835, + "num_input_tokens_seen": 504544, + "step": 1890 + }, + { + "epoch": 0.4967234600262123, + "grad_norm": 3.8820722103118896, + "learning_rate": 2.4823066841415465e-05, + "loss": 0.5434, + "num_input_tokens_seen": 505728, + "step": 1895 + }, + { + "epoch": 0.4980340760157274, + "grad_norm": 31.022531509399414, + "learning_rate": 2.488859764089122e-05, + "loss": 0.6397, + "num_input_tokens_seen": 506640, + "step": 1900 + }, + { + "epoch": 0.49934469200524245, + "grad_norm": 2.1646859645843506, + "learning_rate": 2.4954128440366977e-05, + "loss": 0.4217, + "num_input_tokens_seen": 507904, + "step": 1905 + }, + { + "epoch": 0.5001310615989515, + "eval_loss": 0.5446770191192627, + "eval_runtime": 16.8838, + "eval_samples_per_second": 50.226, + "eval_steps_per_second": 25.113, + "num_input_tokens_seen": 508608, + "step": 1908 + }, + { + "epoch": 0.5006553079947575, + "grad_norm": 3.7143924236297607, + "learning_rate": 2.501965923984273e-05, + "loss": 0.5547, + "num_input_tokens_seen": 508992, + "step": 1910 + }, + { + "epoch": 0.5019659239842726, + "grad_norm": 2.5707626342773438, + "learning_rate": 2.5085190039318478e-05, + "loss": 0.3163, + "num_input_tokens_seen": 510256, + "step": 1915 + }, + { + "epoch": 0.5032765399737876, + "grad_norm": 4.543787479400635, + "learning_rate": 2.5150720838794235e-05, + "loss": 0.5593, + "num_input_tokens_seen": 511536, + "step": 1920 + }, + { + "epoch": 0.5045871559633027, + "grad_norm": 2.77130389213562, + "learning_rate": 2.521625163826999e-05, + "loss": 0.6053, + "num_input_tokens_seen": 512992, + "step": 1925 + }, + { + "epoch": 0.5058977719528178, + "grad_norm": 2.372844934463501, + "learning_rate": 2.5281782437745743e-05, + "loss": 0.4414, + "num_input_tokens_seen": 514208, + "step": 1930 + }, + { + "epoch": 0.5072083879423329, + "grad_norm": 2.5371577739715576, + "learning_rate": 2.5347313237221493e-05, + "loss": 0.7122, + "num_input_tokens_seen": 515536, + "step": 1935 + }, + { + "epoch": 0.508519003931848, + "grad_norm": 1.6292537450790405, + "learning_rate": 2.5412844036697247e-05, + "loss": 0.6212, + "num_input_tokens_seen": 516720, + "step": 1940 + }, + { + "epoch": 0.509829619921363, + "grad_norm": 2.897817850112915, + "learning_rate": 2.5478374836173e-05, + "loss": 0.4091, + "num_input_tokens_seen": 517808, + "step": 1945 + }, + { + "epoch": 0.5111402359108781, + "grad_norm": 2.750302314758301, + "learning_rate": 2.554390563564876e-05, + "loss": 0.4856, + "num_input_tokens_seen": 519312, + "step": 1950 + }, + { + "epoch": 0.5124508519003932, + "grad_norm": 1.7094156742095947, + "learning_rate": 2.5609436435124512e-05, + "loss": 0.4891, + "num_input_tokens_seen": 520528, + "step": 1955 + }, + { + "epoch": 0.5137614678899083, + "grad_norm": 4.833734035491943, + "learning_rate": 2.5674967234600263e-05, + "loss": 0.4792, + "num_input_tokens_seen": 521744, + "step": 1960 + }, + { + "epoch": 0.5150720838794234, + "grad_norm": 8.626018524169922, + "learning_rate": 2.5740498034076017e-05, + "loss": 0.4871, + "num_input_tokens_seen": 522976, + "step": 1965 + }, + { + "epoch": 0.5163826998689384, + "grad_norm": 22.197036743164062, + "learning_rate": 2.580602883355177e-05, + "loss": 0.4985, + "num_input_tokens_seen": 524560, + "step": 1970 + }, + { + "epoch": 0.5176933158584535, + "grad_norm": 1.098995566368103, + "learning_rate": 2.5871559633027525e-05, + "loss": 0.3378, + "num_input_tokens_seen": 525968, + "step": 1975 + }, + { + "epoch": 0.5190039318479686, + "grad_norm": 0.8045446276664734, + "learning_rate": 2.5937090432503282e-05, + "loss": 0.5797, + "num_input_tokens_seen": 527808, + "step": 1980 + }, + { + "epoch": 0.5203145478374837, + "grad_norm": 2.522228240966797, + "learning_rate": 2.600262123197903e-05, + "loss": 0.4691, + "num_input_tokens_seen": 529472, + "step": 1985 + }, + { + "epoch": 0.5216251638269986, + "grad_norm": 4.142117023468018, + "learning_rate": 2.6068152031454783e-05, + "loss": 0.5555, + "num_input_tokens_seen": 530864, + "step": 1990 + }, + { + "epoch": 0.5229357798165137, + "grad_norm": 1.5756789445877075, + "learning_rate": 2.613368283093054e-05, + "loss": 0.5429, + "num_input_tokens_seen": 532448, + "step": 1995 + }, + { + "epoch": 0.5242463958060288, + "grad_norm": 1.192270278930664, + "learning_rate": 2.6199213630406294e-05, + "loss": 0.5112, + "num_input_tokens_seen": 534064, + "step": 2000 + }, + { + "epoch": 0.5255570117955439, + "grad_norm": 4.7171149253845215, + "learning_rate": 2.6264744429882045e-05, + "loss": 0.6287, + "num_input_tokens_seen": 535328, + "step": 2005 + }, + { + "epoch": 0.526867627785059, + "grad_norm": 6.241026401519775, + "learning_rate": 2.63302752293578e-05, + "loss": 0.7761, + "num_input_tokens_seen": 536576, + "step": 2010 + }, + { + "epoch": 0.528178243774574, + "grad_norm": 2.6102192401885986, + "learning_rate": 2.6395806028833552e-05, + "loss": 0.3863, + "num_input_tokens_seen": 537792, + "step": 2015 + }, + { + "epoch": 0.5294888597640891, + "grad_norm": 1.9525785446166992, + "learning_rate": 2.6461336828309306e-05, + "loss": 0.4048, + "num_input_tokens_seen": 539120, + "step": 2020 + }, + { + "epoch": 0.5307994757536042, + "grad_norm": 2.853100061416626, + "learning_rate": 2.6526867627785064e-05, + "loss": 0.5222, + "num_input_tokens_seen": 540320, + "step": 2025 + }, + { + "epoch": 0.5321100917431193, + "grad_norm": 1.9760186672210693, + "learning_rate": 2.659239842726081e-05, + "loss": 0.5513, + "num_input_tokens_seen": 541760, + "step": 2030 + }, + { + "epoch": 0.5334207077326344, + "grad_norm": 9.311144828796387, + "learning_rate": 2.6657929226736568e-05, + "loss": 0.6737, + "num_input_tokens_seen": 542784, + "step": 2035 + }, + { + "epoch": 0.5347313237221494, + "grad_norm": 2.6561102867126465, + "learning_rate": 2.6723460026212322e-05, + "loss": 0.6018, + "num_input_tokens_seen": 543968, + "step": 2040 + }, + { + "epoch": 0.5360419397116645, + "grad_norm": 1.203856110572815, + "learning_rate": 2.6788990825688076e-05, + "loss": 0.4169, + "num_input_tokens_seen": 545248, + "step": 2045 + }, + { + "epoch": 0.5373525557011796, + "grad_norm": 4.357290744781494, + "learning_rate": 2.685452162516383e-05, + "loss": 0.7056, + "num_input_tokens_seen": 546400, + "step": 2050 + }, + { + "epoch": 0.5386631716906947, + "grad_norm": 2.096261978149414, + "learning_rate": 2.692005242463958e-05, + "loss": 0.4569, + "num_input_tokens_seen": 547856, + "step": 2055 + }, + { + "epoch": 0.5399737876802096, + "grad_norm": 1.8115581274032593, + "learning_rate": 2.6985583224115334e-05, + "loss": 0.3972, + "num_input_tokens_seen": 549152, + "step": 2060 + }, + { + "epoch": 0.5412844036697247, + "grad_norm": 8.141670227050781, + "learning_rate": 2.7051114023591088e-05, + "loss": 0.4606, + "num_input_tokens_seen": 550656, + "step": 2065 + }, + { + "epoch": 0.5425950196592398, + "grad_norm": 3.558990001678467, + "learning_rate": 2.7116644823066845e-05, + "loss": 0.4585, + "num_input_tokens_seen": 551680, + "step": 2070 + }, + { + "epoch": 0.5439056356487549, + "grad_norm": 1.4862871170043945, + "learning_rate": 2.7182175622542593e-05, + "loss": 0.4549, + "num_input_tokens_seen": 553088, + "step": 2075 + }, + { + "epoch": 0.54521625163827, + "grad_norm": 4.9191389083862305, + "learning_rate": 2.724770642201835e-05, + "loss": 0.4723, + "num_input_tokens_seen": 554304, + "step": 2080 + }, + { + "epoch": 0.546526867627785, + "grad_norm": 2.5342025756835938, + "learning_rate": 2.7313237221494104e-05, + "loss": 0.6532, + "num_input_tokens_seen": 555584, + "step": 2085 + }, + { + "epoch": 0.5478374836173001, + "grad_norm": 2.670325994491577, + "learning_rate": 2.7378768020969858e-05, + "loss": 0.4908, + "num_input_tokens_seen": 556768, + "step": 2090 + }, + { + "epoch": 0.5491480996068152, + "grad_norm": 2.1538589000701904, + "learning_rate": 2.744429882044561e-05, + "loss": 0.4112, + "num_input_tokens_seen": 558208, + "step": 2095 + }, + { + "epoch": 0.5504587155963303, + "grad_norm": 2.252680778503418, + "learning_rate": 2.7509829619921362e-05, + "loss": 0.5277, + "num_input_tokens_seen": 559664, + "step": 2100 + }, + { + "epoch": 0.5517693315858454, + "grad_norm": 3.034187078475952, + "learning_rate": 2.7575360419397116e-05, + "loss": 0.5098, + "num_input_tokens_seen": 560800, + "step": 2105 + }, + { + "epoch": 0.5530799475753604, + "grad_norm": 2.8518550395965576, + "learning_rate": 2.7640891218872873e-05, + "loss": 0.521, + "num_input_tokens_seen": 562640, + "step": 2110 + }, + { + "epoch": 0.5543905635648755, + "grad_norm": 1.8912276029586792, + "learning_rate": 2.7706422018348627e-05, + "loss": 0.5802, + "num_input_tokens_seen": 563760, + "step": 2115 + }, + { + "epoch": 0.5557011795543906, + "grad_norm": 5.148646831512451, + "learning_rate": 2.777195281782438e-05, + "loss": 0.4479, + "num_input_tokens_seen": 565280, + "step": 2120 + }, + { + "epoch": 0.5570117955439057, + "grad_norm": 1.9108638763427734, + "learning_rate": 2.7837483617300132e-05, + "loss": 0.4951, + "num_input_tokens_seen": 566560, + "step": 2125 + }, + { + "epoch": 0.5583224115334207, + "grad_norm": 40.52590560913086, + "learning_rate": 2.7903014416775886e-05, + "loss": 0.3798, + "num_input_tokens_seen": 567776, + "step": 2130 + }, + { + "epoch": 0.5596330275229358, + "grad_norm": 5.895997047424316, + "learning_rate": 2.796854521625164e-05, + "loss": 0.6738, + "num_input_tokens_seen": 568960, + "step": 2135 + }, + { + "epoch": 0.5609436435124509, + "grad_norm": 1.3565953969955444, + "learning_rate": 2.8034076015727394e-05, + "loss": 0.5387, + "num_input_tokens_seen": 570432, + "step": 2140 + }, + { + "epoch": 0.562254259501966, + "grad_norm": 1.972806453704834, + "learning_rate": 2.8099606815203144e-05, + "loss": 0.5822, + "num_input_tokens_seen": 571408, + "step": 2145 + }, + { + "epoch": 0.563564875491481, + "grad_norm": 3.1153697967529297, + "learning_rate": 2.8165137614678898e-05, + "loss": 0.508, + "num_input_tokens_seen": 572592, + "step": 2150 + }, + { + "epoch": 0.564875491480996, + "grad_norm": 1.3652009963989258, + "learning_rate": 2.8230668414154655e-05, + "loss": 0.5673, + "num_input_tokens_seen": 573968, + "step": 2155 + }, + { + "epoch": 0.5661861074705111, + "grad_norm": 7.365413665771484, + "learning_rate": 2.829619921363041e-05, + "loss": 0.5048, + "num_input_tokens_seen": 575088, + "step": 2160 + }, + { + "epoch": 0.5674967234600262, + "grad_norm": 15.578024864196777, + "learning_rate": 2.8361730013106163e-05, + "loss": 0.5669, + "num_input_tokens_seen": 576368, + "step": 2165 + }, + { + "epoch": 0.5688073394495413, + "grad_norm": 5.52711820602417, + "learning_rate": 2.8427260812581914e-05, + "loss": 0.6036, + "num_input_tokens_seen": 577520, + "step": 2170 + }, + { + "epoch": 0.5701179554390564, + "grad_norm": 2.842482566833496, + "learning_rate": 2.8492791612057668e-05, + "loss": 0.7618, + "num_input_tokens_seen": 579152, + "step": 2175 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 2.0068747997283936, + "learning_rate": 2.855832241153342e-05, + "loss": 0.3917, + "num_input_tokens_seen": 580720, + "step": 2180 + }, + { + "epoch": 0.5727391874180865, + "grad_norm": 3.681387186050415, + "learning_rate": 2.862385321100918e-05, + "loss": 0.4119, + "num_input_tokens_seen": 582128, + "step": 2185 + }, + { + "epoch": 0.5740498034076016, + "grad_norm": 3.3330817222595215, + "learning_rate": 2.8689384010484926e-05, + "loss": 0.4981, + "num_input_tokens_seen": 584000, + "step": 2190 + }, + { + "epoch": 0.5753604193971167, + "grad_norm": 5.041707992553711, + "learning_rate": 2.875491480996068e-05, + "loss": 0.7631, + "num_input_tokens_seen": 585104, + "step": 2195 + }, + { + "epoch": 0.5766710353866317, + "grad_norm": 1.892556071281433, + "learning_rate": 2.8820445609436437e-05, + "loss": 0.4555, + "num_input_tokens_seen": 586432, + "step": 2200 + }, + { + "epoch": 0.5779816513761468, + "grad_norm": 1.6688127517700195, + "learning_rate": 2.888597640891219e-05, + "loss": 0.7242, + "num_input_tokens_seen": 587648, + "step": 2205 + }, + { + "epoch": 0.5792922673656619, + "grad_norm": 5.751414775848389, + "learning_rate": 2.8951507208387945e-05, + "loss": 0.5581, + "num_input_tokens_seen": 588672, + "step": 2210 + }, + { + "epoch": 0.580602883355177, + "grad_norm": 2.6093523502349854, + "learning_rate": 2.9017038007863695e-05, + "loss": 0.4202, + "num_input_tokens_seen": 590192, + "step": 2215 + }, + { + "epoch": 0.581913499344692, + "grad_norm": 4.827092170715332, + "learning_rate": 2.908256880733945e-05, + "loss": 0.5273, + "num_input_tokens_seen": 591072, + "step": 2220 + }, + { + "epoch": 0.583224115334207, + "grad_norm": 1.4871724843978882, + "learning_rate": 2.9148099606815203e-05, + "loss": 0.3957, + "num_input_tokens_seen": 592192, + "step": 2225 + }, + { + "epoch": 0.5845347313237221, + "grad_norm": 2.08154034614563, + "learning_rate": 2.921363040629096e-05, + "loss": 0.5306, + "num_input_tokens_seen": 593520, + "step": 2230 + }, + { + "epoch": 0.5858453473132372, + "grad_norm": 4.367341995239258, + "learning_rate": 2.9279161205766714e-05, + "loss": 0.4037, + "num_input_tokens_seen": 594528, + "step": 2235 + }, + { + "epoch": 0.5871559633027523, + "grad_norm": 1.4721276760101318, + "learning_rate": 2.9344692005242465e-05, + "loss": 0.388, + "num_input_tokens_seen": 595824, + "step": 2240 + }, + { + "epoch": 0.5884665792922673, + "grad_norm": 2.1789708137512207, + "learning_rate": 2.941022280471822e-05, + "loss": 0.3229, + "num_input_tokens_seen": 597136, + "step": 2245 + }, + { + "epoch": 0.5897771952817824, + "grad_norm": 10.080060005187988, + "learning_rate": 2.9475753604193973e-05, + "loss": 0.591, + "num_input_tokens_seen": 598320, + "step": 2250 + }, + { + "epoch": 0.5910878112712975, + "grad_norm": 2.7720048427581787, + "learning_rate": 2.9541284403669727e-05, + "loss": 0.4191, + "num_input_tokens_seen": 599792, + "step": 2255 + }, + { + "epoch": 0.5923984272608126, + "grad_norm": 2.0886340141296387, + "learning_rate": 2.9606815203145477e-05, + "loss": 0.3517, + "num_input_tokens_seen": 601120, + "step": 2260 + }, + { + "epoch": 0.5937090432503277, + "grad_norm": 1.4340503215789795, + "learning_rate": 2.967234600262123e-05, + "loss": 0.4616, + "num_input_tokens_seen": 602368, + "step": 2265 + }, + { + "epoch": 0.5950196592398427, + "grad_norm": 1.232086181640625, + "learning_rate": 2.9737876802096985e-05, + "loss": 0.358, + "num_input_tokens_seen": 603824, + "step": 2270 + }, + { + "epoch": 0.5963302752293578, + "grad_norm": 2.3230905532836914, + "learning_rate": 2.9803407601572742e-05, + "loss": 0.5391, + "num_input_tokens_seen": 605376, + "step": 2275 + }, + { + "epoch": 0.5976408912188729, + "grad_norm": 7.186913967132568, + "learning_rate": 2.9868938401048496e-05, + "loss": 0.5157, + "num_input_tokens_seen": 606672, + "step": 2280 + }, + { + "epoch": 0.598951507208388, + "grad_norm": 1.3951942920684814, + "learning_rate": 2.9934469200524247e-05, + "loss": 0.3326, + "num_input_tokens_seen": 608096, + "step": 2285 + }, + { + "epoch": 0.6002621231979031, + "grad_norm": 1.6297417879104614, + "learning_rate": 3e-05, + "loss": 0.3878, + "num_input_tokens_seen": 609408, + "step": 2290 + }, + { + "epoch": 0.601572739187418, + "grad_norm": 2.162332773208618, + "learning_rate": 3.0065530799475755e-05, + "loss": 0.4968, + "num_input_tokens_seen": 610720, + "step": 2295 + }, + { + "epoch": 0.6028833551769331, + "grad_norm": 2.4945149421691895, + "learning_rate": 3.013106159895151e-05, + "loss": 0.6538, + "num_input_tokens_seen": 611888, + "step": 2300 + }, + { + "epoch": 0.6041939711664482, + "grad_norm": 1.2398974895477295, + "learning_rate": 3.0196592398427266e-05, + "loss": 0.6401, + "num_input_tokens_seen": 613008, + "step": 2305 + }, + { + "epoch": 0.6055045871559633, + "grad_norm": 2.5096116065979004, + "learning_rate": 3.0262123197903013e-05, + "loss": 0.4298, + "num_input_tokens_seen": 614096, + "step": 2310 + }, + { + "epoch": 0.6068152031454783, + "grad_norm": 2.3301234245300293, + "learning_rate": 3.032765399737877e-05, + "loss": 0.5405, + "num_input_tokens_seen": 615088, + "step": 2315 + }, + { + "epoch": 0.6081258191349934, + "grad_norm": 5.256295204162598, + "learning_rate": 3.0393184796854524e-05, + "loss": 0.5642, + "num_input_tokens_seen": 616400, + "step": 2320 + }, + { + "epoch": 0.6094364351245085, + "grad_norm": 1.6257212162017822, + "learning_rate": 3.0458715596330278e-05, + "loss": 0.6367, + "num_input_tokens_seen": 618032, + "step": 2325 + }, + { + "epoch": 0.6107470511140236, + "grad_norm": 2.639606237411499, + "learning_rate": 3.052424639580603e-05, + "loss": 0.6793, + "num_input_tokens_seen": 619296, + "step": 2330 + }, + { + "epoch": 0.6120576671035387, + "grad_norm": 3.0963141918182373, + "learning_rate": 3.058977719528178e-05, + "loss": 0.4537, + "num_input_tokens_seen": 620736, + "step": 2335 + }, + { + "epoch": 0.6133682830930537, + "grad_norm": 3.3922524452209473, + "learning_rate": 3.0655307994757537e-05, + "loss": 0.7121, + "num_input_tokens_seen": 621888, + "step": 2340 + }, + { + "epoch": 0.6146788990825688, + "grad_norm": 3.1200520992279053, + "learning_rate": 3.0720838794233294e-05, + "loss": 0.4692, + "num_input_tokens_seen": 622992, + "step": 2345 + }, + { + "epoch": 0.6159895150720839, + "grad_norm": 0.29429489374160767, + "learning_rate": 3.0786369593709044e-05, + "loss": 0.6501, + "num_input_tokens_seen": 624304, + "step": 2350 + }, + { + "epoch": 0.617300131061599, + "grad_norm": 1.4841248989105225, + "learning_rate": 3.0851900393184795e-05, + "loss": 0.5157, + "num_input_tokens_seen": 625744, + "step": 2355 + }, + { + "epoch": 0.6186107470511141, + "grad_norm": 2.1199328899383545, + "learning_rate": 3.091743119266055e-05, + "loss": 0.5637, + "num_input_tokens_seen": 627120, + "step": 2360 + }, + { + "epoch": 0.6199213630406291, + "grad_norm": 4.293496131896973, + "learning_rate": 3.09829619921363e-05, + "loss": 0.6202, + "num_input_tokens_seen": 628144, + "step": 2365 + }, + { + "epoch": 0.6212319790301442, + "grad_norm": 1.6403164863586426, + "learning_rate": 3.104849279161206e-05, + "loss": 0.3705, + "num_input_tokens_seen": 629344, + "step": 2370 + }, + { + "epoch": 0.6225425950196593, + "grad_norm": 6.029444694519043, + "learning_rate": 3.111402359108782e-05, + "loss": 0.7921, + "num_input_tokens_seen": 630928, + "step": 2375 + }, + { + "epoch": 0.6238532110091743, + "grad_norm": 1.8292337656021118, + "learning_rate": 3.117955439056357e-05, + "loss": 0.5241, + "num_input_tokens_seen": 632272, + "step": 2380 + }, + { + "epoch": 0.6251638269986893, + "grad_norm": 1.1080173254013062, + "learning_rate": 3.124508519003932e-05, + "loss": 0.5289, + "num_input_tokens_seen": 633440, + "step": 2385 + }, + { + "epoch": 0.6264744429882044, + "grad_norm": 2.6049928665161133, + "learning_rate": 3.1310615989515076e-05, + "loss": 0.4768, + "num_input_tokens_seen": 634624, + "step": 2390 + }, + { + "epoch": 0.6277850589777195, + "grad_norm": 1.507249116897583, + "learning_rate": 3.1376146788990826e-05, + "loss": 0.8147, + "num_input_tokens_seen": 635664, + "step": 2395 + }, + { + "epoch": 0.6290956749672346, + "grad_norm": 1.5870813131332397, + "learning_rate": 3.144167758846658e-05, + "loss": 0.4665, + "num_input_tokens_seen": 637104, + "step": 2400 + }, + { + "epoch": 0.6304062909567497, + "grad_norm": 0.9337738156318665, + "learning_rate": 3.1507208387942334e-05, + "loss": 0.3475, + "num_input_tokens_seen": 638448, + "step": 2405 + }, + { + "epoch": 0.6317169069462647, + "grad_norm": 2.454340934753418, + "learning_rate": 3.1572739187418085e-05, + "loss": 0.3581, + "num_input_tokens_seen": 641536, + "step": 2410 + }, + { + "epoch": 0.6330275229357798, + "grad_norm": 1.0215712785720825, + "learning_rate": 3.163826998689384e-05, + "loss": 0.4593, + "num_input_tokens_seen": 642960, + "step": 2415 + }, + { + "epoch": 0.6343381389252949, + "grad_norm": 3.1397125720977783, + "learning_rate": 3.17038007863696e-05, + "loss": 0.5151, + "num_input_tokens_seen": 643952, + "step": 2420 + }, + { + "epoch": 0.63564875491481, + "grad_norm": 4.557147026062012, + "learning_rate": 3.176933158584535e-05, + "loss": 0.5031, + "num_input_tokens_seen": 645184, + "step": 2425 + }, + { + "epoch": 0.6369593709043251, + "grad_norm": 2.932995319366455, + "learning_rate": 3.18348623853211e-05, + "loss": 0.6049, + "num_input_tokens_seen": 646880, + "step": 2430 + }, + { + "epoch": 0.6382699868938401, + "grad_norm": 1.8398240804672241, + "learning_rate": 3.190039318479686e-05, + "loss": 0.4255, + "num_input_tokens_seen": 648064, + "step": 2435 + }, + { + "epoch": 0.6395806028833552, + "grad_norm": 2.2856993675231934, + "learning_rate": 3.196592398427261e-05, + "loss": 0.72, + "num_input_tokens_seen": 649552, + "step": 2440 + }, + { + "epoch": 0.6408912188728703, + "grad_norm": 1.9679455757141113, + "learning_rate": 3.2031454783748365e-05, + "loss": 0.5957, + "num_input_tokens_seen": 650848, + "step": 2445 + }, + { + "epoch": 0.6422018348623854, + "grad_norm": 2.0191407203674316, + "learning_rate": 3.2096985583224116e-05, + "loss": 0.7229, + "num_input_tokens_seen": 652240, + "step": 2450 + }, + { + "epoch": 0.6435124508519003, + "grad_norm": 2.110394239425659, + "learning_rate": 3.216251638269987e-05, + "loss": 0.4842, + "num_input_tokens_seen": 653616, + "step": 2455 + }, + { + "epoch": 0.6448230668414154, + "grad_norm": 17.019651412963867, + "learning_rate": 3.2228047182175624e-05, + "loss": 0.7283, + "num_input_tokens_seen": 654864, + "step": 2460 + }, + { + "epoch": 0.6461336828309305, + "grad_norm": 3.0046138763427734, + "learning_rate": 3.229357798165138e-05, + "loss": 0.3973, + "num_input_tokens_seen": 656288, + "step": 2465 + }, + { + "epoch": 0.6474442988204456, + "grad_norm": 7.582738876342773, + "learning_rate": 3.235910878112713e-05, + "loss": 0.7185, + "num_input_tokens_seen": 657360, + "step": 2470 + }, + { + "epoch": 0.6487549148099607, + "grad_norm": 2.482320547103882, + "learning_rate": 3.242463958060288e-05, + "loss": 0.6217, + "num_input_tokens_seen": 658704, + "step": 2475 + }, + { + "epoch": 0.6500655307994757, + "grad_norm": 7.649007320404053, + "learning_rate": 3.249017038007864e-05, + "loss": 0.5055, + "num_input_tokens_seen": 659792, + "step": 2480 + }, + { + "epoch": 0.6513761467889908, + "grad_norm": 2.968649387359619, + "learning_rate": 3.255570117955439e-05, + "loss": 0.4254, + "num_input_tokens_seen": 661152, + "step": 2485 + }, + { + "epoch": 0.6526867627785059, + "grad_norm": 1.8702565431594849, + "learning_rate": 3.262123197903015e-05, + "loss": 0.3319, + "num_input_tokens_seen": 662288, + "step": 2490 + }, + { + "epoch": 0.653997378768021, + "grad_norm": 1.7817875146865845, + "learning_rate": 3.26867627785059e-05, + "loss": 0.5574, + "num_input_tokens_seen": 663568, + "step": 2495 + }, + { + "epoch": 0.6553079947575361, + "grad_norm": 4.5275397300720215, + "learning_rate": 3.2752293577981655e-05, + "loss": 0.5624, + "num_input_tokens_seen": 664960, + "step": 2500 + }, + { + "epoch": 0.6566186107470511, + "grad_norm": 2.402841091156006, + "learning_rate": 3.2817824377457405e-05, + "loss": 0.4615, + "num_input_tokens_seen": 666416, + "step": 2505 + }, + { + "epoch": 0.6579292267365662, + "grad_norm": 4.630550861358643, + "learning_rate": 3.288335517693316e-05, + "loss": 0.6696, + "num_input_tokens_seen": 667488, + "step": 2510 + }, + { + "epoch": 0.6592398427260813, + "grad_norm": 1.484650731086731, + "learning_rate": 3.294888597640891e-05, + "loss": 0.2681, + "num_input_tokens_seen": 669072, + "step": 2515 + }, + { + "epoch": 0.6605504587155964, + "grad_norm": 2.1597440242767334, + "learning_rate": 3.3014416775884664e-05, + "loss": 0.6086, + "num_input_tokens_seen": 670256, + "step": 2520 + }, + { + "epoch": 0.6618610747051114, + "grad_norm": 5.3292555809021, + "learning_rate": 3.307994757536042e-05, + "loss": 0.4122, + "num_input_tokens_seen": 671520, + "step": 2525 + }, + { + "epoch": 0.6631716906946264, + "grad_norm": 4.173333644866943, + "learning_rate": 3.314547837483618e-05, + "loss": 0.605, + "num_input_tokens_seen": 673056, + "step": 2530 + }, + { + "epoch": 0.6644823066841415, + "grad_norm": 1.3014944791793823, + "learning_rate": 3.321100917431193e-05, + "loss": 0.4572, + "num_input_tokens_seen": 674096, + "step": 2535 + }, + { + "epoch": 0.6657929226736566, + "grad_norm": 1.481396198272705, + "learning_rate": 3.327653997378768e-05, + "loss": 0.3817, + "num_input_tokens_seen": 675264, + "step": 2540 + }, + { + "epoch": 0.6671035386631717, + "grad_norm": 2.1976869106292725, + "learning_rate": 3.334207077326344e-05, + "loss": 0.4447, + "num_input_tokens_seen": 676784, + "step": 2545 + }, + { + "epoch": 0.6684141546526867, + "grad_norm": 10.321023941040039, + "learning_rate": 3.340760157273919e-05, + "loss": 0.5841, + "num_input_tokens_seen": 678064, + "step": 2550 + }, + { + "epoch": 0.6697247706422018, + "grad_norm": 1.4158951044082642, + "learning_rate": 3.3473132372214945e-05, + "loss": 0.5001, + "num_input_tokens_seen": 679728, + "step": 2555 + }, + { + "epoch": 0.6710353866317169, + "grad_norm": 0.9874671697616577, + "learning_rate": 3.3538663171690695e-05, + "loss": 0.4693, + "num_input_tokens_seen": 680976, + "step": 2560 + }, + { + "epoch": 0.672346002621232, + "grad_norm": 1.0344654321670532, + "learning_rate": 3.3604193971166446e-05, + "loss": 0.421, + "num_input_tokens_seen": 682512, + "step": 2565 + }, + { + "epoch": 0.6736566186107471, + "grad_norm": 1.5576914548873901, + "learning_rate": 3.36697247706422e-05, + "loss": 0.5373, + "num_input_tokens_seen": 684080, + "step": 2570 + }, + { + "epoch": 0.6749672346002621, + "grad_norm": 1.646523356437683, + "learning_rate": 3.373525557011796e-05, + "loss": 0.4022, + "num_input_tokens_seen": 685248, + "step": 2575 + }, + { + "epoch": 0.6762778505897772, + "grad_norm": 1.1741948127746582, + "learning_rate": 3.380078636959371e-05, + "loss": 0.5071, + "num_input_tokens_seen": 686480, + "step": 2580 + }, + { + "epoch": 0.6775884665792923, + "grad_norm": 2.974672317504883, + "learning_rate": 3.386631716906947e-05, + "loss": 0.5378, + "num_input_tokens_seen": 687760, + "step": 2585 + }, + { + "epoch": 0.6788990825688074, + "grad_norm": 2.0403947830200195, + "learning_rate": 3.393184796854522e-05, + "loss": 0.5509, + "num_input_tokens_seen": 688896, + "step": 2590 + }, + { + "epoch": 0.6802096985583224, + "grad_norm": 1.5459450483322144, + "learning_rate": 3.399737876802097e-05, + "loss": 0.5407, + "num_input_tokens_seen": 689968, + "step": 2595 + }, + { + "epoch": 0.6815203145478375, + "grad_norm": 1.4326729774475098, + "learning_rate": 3.4062909567496726e-05, + "loss": 0.5807, + "num_input_tokens_seen": 691232, + "step": 2600 + }, + { + "epoch": 0.6828309305373526, + "grad_norm": 1.5758345127105713, + "learning_rate": 3.4128440366972484e-05, + "loss": 0.801, + "num_input_tokens_seen": 692304, + "step": 2605 + }, + { + "epoch": 0.6841415465268676, + "grad_norm": 0.7562657594680786, + "learning_rate": 3.419397116644823e-05, + "loss": 0.4031, + "num_input_tokens_seen": 693744, + "step": 2610 + }, + { + "epoch": 0.6854521625163827, + "grad_norm": 2.1749532222747803, + "learning_rate": 3.4259501965923985e-05, + "loss": 0.6148, + "num_input_tokens_seen": 695184, + "step": 2615 + }, + { + "epoch": 0.6867627785058977, + "grad_norm": 1.6033637523651123, + "learning_rate": 3.432503276539974e-05, + "loss": 0.6659, + "num_input_tokens_seen": 696416, + "step": 2620 + }, + { + "epoch": 0.6880733944954128, + "grad_norm": 1.316315770149231, + "learning_rate": 3.439056356487549e-05, + "loss": 0.4785, + "num_input_tokens_seen": 697824, + "step": 2625 + }, + { + "epoch": 0.6893840104849279, + "grad_norm": 0.9042149782180786, + "learning_rate": 3.445609436435125e-05, + "loss": 0.5656, + "num_input_tokens_seen": 699328, + "step": 2630 + }, + { + "epoch": 0.690694626474443, + "grad_norm": 1.6404377222061157, + "learning_rate": 3.4521625163827e-05, + "loss": 0.583, + "num_input_tokens_seen": 700320, + "step": 2635 + }, + { + "epoch": 0.6920052424639581, + "grad_norm": 1.0682393312454224, + "learning_rate": 3.458715596330275e-05, + "loss": 0.5581, + "num_input_tokens_seen": 701712, + "step": 2640 + }, + { + "epoch": 0.6933158584534731, + "grad_norm": 5.412893772125244, + "learning_rate": 3.465268676277851e-05, + "loss": 0.5892, + "num_input_tokens_seen": 703952, + "step": 2645 + }, + { + "epoch": 0.6946264744429882, + "grad_norm": 0.766356348991394, + "learning_rate": 3.4718217562254266e-05, + "loss": 0.4913, + "num_input_tokens_seen": 705168, + "step": 2650 + }, + { + "epoch": 0.6959370904325033, + "grad_norm": 2.2321724891662598, + "learning_rate": 3.4783748361730016e-05, + "loss": 0.3102, + "num_input_tokens_seen": 706272, + "step": 2655 + }, + { + "epoch": 0.6972477064220184, + "grad_norm": 1.082716941833496, + "learning_rate": 3.484927916120577e-05, + "loss": 0.3813, + "num_input_tokens_seen": 707680, + "step": 2660 + }, + { + "epoch": 0.6985583224115334, + "grad_norm": 2.015444278717041, + "learning_rate": 3.4914809960681524e-05, + "loss": 0.515, + "num_input_tokens_seen": 709024, + "step": 2665 + }, + { + "epoch": 0.6998689384010485, + "grad_norm": 3.4212193489074707, + "learning_rate": 3.4980340760157274e-05, + "loss": 0.491, + "num_input_tokens_seen": 710576, + "step": 2670 + }, + { + "epoch": 0.7011795543905636, + "grad_norm": 1.334208607673645, + "learning_rate": 3.504587155963303e-05, + "loss": 0.7212, + "num_input_tokens_seen": 711984, + "step": 2675 + }, + { + "epoch": 0.7024901703800787, + "grad_norm": 4.645016193389893, + "learning_rate": 3.511140235910878e-05, + "loss": 0.4074, + "num_input_tokens_seen": 713344, + "step": 2680 + }, + { + "epoch": 0.7038007863695938, + "grad_norm": 0.8785014748573303, + "learning_rate": 3.517693315858453e-05, + "loss": 0.405, + "num_input_tokens_seen": 714704, + "step": 2685 + }, + { + "epoch": 0.7051114023591087, + "grad_norm": 2.475933074951172, + "learning_rate": 3.524246395806029e-05, + "loss": 0.6673, + "num_input_tokens_seen": 716288, + "step": 2690 + }, + { + "epoch": 0.7064220183486238, + "grad_norm": 2.372718095779419, + "learning_rate": 3.530799475753605e-05, + "loss": 0.4694, + "num_input_tokens_seen": 717392, + "step": 2695 + }, + { + "epoch": 0.7077326343381389, + "grad_norm": 5.733434200286865, + "learning_rate": 3.53735255570118e-05, + "loss": 0.6381, + "num_input_tokens_seen": 718784, + "step": 2700 + }, + { + "epoch": 0.709043250327654, + "grad_norm": 2.7929539680480957, + "learning_rate": 3.543905635648755e-05, + "loss": 0.5704, + "num_input_tokens_seen": 719824, + "step": 2705 + }, + { + "epoch": 0.7103538663171691, + "grad_norm": 1.0463610887527466, + "learning_rate": 3.5504587155963306e-05, + "loss": 0.4701, + "num_input_tokens_seen": 721424, + "step": 2710 + }, + { + "epoch": 0.7116644823066841, + "grad_norm": 1.7203569412231445, + "learning_rate": 3.5570117955439056e-05, + "loss": 0.525, + "num_input_tokens_seen": 722784, + "step": 2715 + }, + { + "epoch": 0.7129750982961992, + "grad_norm": 1.6268484592437744, + "learning_rate": 3.5635648754914814e-05, + "loss": 0.5735, + "num_input_tokens_seen": 724160, + "step": 2720 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 2.5614984035491943, + "learning_rate": 3.570117955439057e-05, + "loss": 0.4989, + "num_input_tokens_seen": 725360, + "step": 2725 + }, + { + "epoch": 0.7155963302752294, + "grad_norm": 2.1210217475891113, + "learning_rate": 3.5766710353866315e-05, + "loss": 0.5118, + "num_input_tokens_seen": 726704, + "step": 2730 + }, + { + "epoch": 0.7169069462647444, + "grad_norm": 0.9489111304283142, + "learning_rate": 3.583224115334207e-05, + "loss": 0.5223, + "num_input_tokens_seen": 728528, + "step": 2735 + }, + { + "epoch": 0.7182175622542595, + "grad_norm": 3.247056245803833, + "learning_rate": 3.589777195281783e-05, + "loss": 0.4888, + "num_input_tokens_seen": 730096, + "step": 2740 + }, + { + "epoch": 0.7195281782437746, + "grad_norm": 2.4805283546447754, + "learning_rate": 3.596330275229358e-05, + "loss": 0.5738, + "num_input_tokens_seen": 731168, + "step": 2745 + }, + { + "epoch": 0.7208387942332897, + "grad_norm": 1.7144755125045776, + "learning_rate": 3.602883355176933e-05, + "loss": 0.4614, + "num_input_tokens_seen": 733136, + "step": 2750 + }, + { + "epoch": 0.7221494102228048, + "grad_norm": 1.199745535850525, + "learning_rate": 3.609436435124509e-05, + "loss": 0.6927, + "num_input_tokens_seen": 734448, + "step": 2755 + }, + { + "epoch": 0.7234600262123198, + "grad_norm": 4.680935382843018, + "learning_rate": 3.615989515072084e-05, + "loss": 0.3673, + "num_input_tokens_seen": 735952, + "step": 2760 + }, + { + "epoch": 0.7247706422018348, + "grad_norm": 1.4281232357025146, + "learning_rate": 3.6225425950196595e-05, + "loss": 0.7013, + "num_input_tokens_seen": 737392, + "step": 2765 + }, + { + "epoch": 0.7260812581913499, + "grad_norm": 4.9637956619262695, + "learning_rate": 3.629095674967235e-05, + "loss": 0.4912, + "num_input_tokens_seen": 738816, + "step": 2770 + }, + { + "epoch": 0.727391874180865, + "grad_norm": 1.4367125034332275, + "learning_rate": 3.6356487549148097e-05, + "loss": 0.5993, + "num_input_tokens_seen": 740336, + "step": 2775 + }, + { + "epoch": 0.72870249017038, + "grad_norm": 3.1905012130737305, + "learning_rate": 3.6422018348623854e-05, + "loss": 0.4276, + "num_input_tokens_seen": 741536, + "step": 2780 + }, + { + "epoch": 0.7300131061598951, + "grad_norm": 2.7692675590515137, + "learning_rate": 3.648754914809961e-05, + "loss": 0.4149, + "num_input_tokens_seen": 743392, + "step": 2785 + }, + { + "epoch": 0.7313237221494102, + "grad_norm": 6.344692230224609, + "learning_rate": 3.655307994757536e-05, + "loss": 0.5001, + "num_input_tokens_seen": 744608, + "step": 2790 + }, + { + "epoch": 0.7326343381389253, + "grad_norm": 9.624946594238281, + "learning_rate": 3.661861074705111e-05, + "loss": 0.4798, + "num_input_tokens_seen": 745840, + "step": 2795 + }, + { + "epoch": 0.7339449541284404, + "grad_norm": 2.40297269821167, + "learning_rate": 3.668414154652687e-05, + "loss": 0.6114, + "num_input_tokens_seen": 746896, + "step": 2800 + }, + { + "epoch": 0.7352555701179554, + "grad_norm": 1.4229652881622314, + "learning_rate": 3.674967234600262e-05, + "loss": 0.6293, + "num_input_tokens_seen": 748176, + "step": 2805 + }, + { + "epoch": 0.7365661861074705, + "grad_norm": 3.364528179168701, + "learning_rate": 3.681520314547838e-05, + "loss": 0.6483, + "num_input_tokens_seen": 749312, + "step": 2810 + }, + { + "epoch": 0.7378768020969856, + "grad_norm": 2.659975051879883, + "learning_rate": 3.6880733944954135e-05, + "loss": 0.5406, + "num_input_tokens_seen": 750864, + "step": 2815 + }, + { + "epoch": 0.7391874180865007, + "grad_norm": 4.2675299644470215, + "learning_rate": 3.694626474442988e-05, + "loss": 0.574, + "num_input_tokens_seen": 751952, + "step": 2820 + }, + { + "epoch": 0.7404980340760158, + "grad_norm": 1.6721526384353638, + "learning_rate": 3.7011795543905636e-05, + "loss": 0.5432, + "num_input_tokens_seen": 752928, + "step": 2825 + }, + { + "epoch": 0.7418086500655308, + "grad_norm": 1.2564460039138794, + "learning_rate": 3.707732634338139e-05, + "loss": 0.747, + "num_input_tokens_seen": 754400, + "step": 2830 + }, + { + "epoch": 0.7431192660550459, + "grad_norm": 1.8826851844787598, + "learning_rate": 3.7142857142857143e-05, + "loss": 0.422, + "num_input_tokens_seen": 755712, + "step": 2835 + }, + { + "epoch": 0.744429882044561, + "grad_norm": 4.330333709716797, + "learning_rate": 3.72083879423329e-05, + "loss": 0.6009, + "num_input_tokens_seen": 758464, + "step": 2840 + }, + { + "epoch": 0.745740498034076, + "grad_norm": 1.775100588798523, + "learning_rate": 3.727391874180865e-05, + "loss": 0.5096, + "num_input_tokens_seen": 759904, + "step": 2845 + }, + { + "epoch": 0.747051114023591, + "grad_norm": 1.440652847290039, + "learning_rate": 3.73394495412844e-05, + "loss": 0.4095, + "num_input_tokens_seen": 761328, + "step": 2850 + }, + { + "epoch": 0.7483617300131061, + "grad_norm": 1.671201229095459, + "learning_rate": 3.740498034076016e-05, + "loss": 0.6293, + "num_input_tokens_seen": 762688, + "step": 2855 + }, + { + "epoch": 0.7496723460026212, + "grad_norm": 2.7217066287994385, + "learning_rate": 3.7470511140235916e-05, + "loss": 0.391, + "num_input_tokens_seen": 763888, + "step": 2860 + }, + { + "epoch": 0.7509829619921363, + "grad_norm": 2.261258602142334, + "learning_rate": 3.753604193971167e-05, + "loss": 0.4864, + "num_input_tokens_seen": 765296, + "step": 2865 + }, + { + "epoch": 0.7522935779816514, + "grad_norm": 1.303938865661621, + "learning_rate": 3.760157273918742e-05, + "loss": 0.5354, + "num_input_tokens_seen": 766928, + "step": 2870 + }, + { + "epoch": 0.7536041939711664, + "grad_norm": 5.035316467285156, + "learning_rate": 3.7667103538663175e-05, + "loss": 0.6004, + "num_input_tokens_seen": 768480, + "step": 2875 + }, + { + "epoch": 0.7549148099606815, + "grad_norm": 18.579681396484375, + "learning_rate": 3.7732634338138925e-05, + "loss": 0.5114, + "num_input_tokens_seen": 769584, + "step": 2880 + }, + { + "epoch": 0.7562254259501966, + "grad_norm": 1.4191893339157104, + "learning_rate": 3.779816513761468e-05, + "loss": 0.5689, + "num_input_tokens_seen": 770960, + "step": 2885 + }, + { + "epoch": 0.7575360419397117, + "grad_norm": 8.866793632507324, + "learning_rate": 3.786369593709043e-05, + "loss": 0.6082, + "num_input_tokens_seen": 772272, + "step": 2890 + }, + { + "epoch": 0.7588466579292268, + "grad_norm": 1.190796136856079, + "learning_rate": 3.7929226736566184e-05, + "loss": 0.3947, + "num_input_tokens_seen": 775536, + "step": 2895 + }, + { + "epoch": 0.7601572739187418, + "grad_norm": 4.693642616271973, + "learning_rate": 3.799475753604194e-05, + "loss": 0.6079, + "num_input_tokens_seen": 777216, + "step": 2900 + }, + { + "epoch": 0.7614678899082569, + "grad_norm": 1.2803094387054443, + "learning_rate": 3.80602883355177e-05, + "loss": 0.4302, + "num_input_tokens_seen": 778320, + "step": 2905 + }, + { + "epoch": 0.762778505897772, + "grad_norm": 1.341591715812683, + "learning_rate": 3.812581913499345e-05, + "loss": 0.5422, + "num_input_tokens_seen": 779712, + "step": 2910 + }, + { + "epoch": 0.7640891218872871, + "grad_norm": 2.90299391746521, + "learning_rate": 3.81913499344692e-05, + "loss": 0.4855, + "num_input_tokens_seen": 780928, + "step": 2915 + }, + { + "epoch": 0.765399737876802, + "grad_norm": 1.2940800189971924, + "learning_rate": 3.825688073394496e-05, + "loss": 0.3741, + "num_input_tokens_seen": 782064, + "step": 2920 + }, + { + "epoch": 0.7667103538663171, + "grad_norm": 1.9918644428253174, + "learning_rate": 3.832241153342071e-05, + "loss": 0.5534, + "num_input_tokens_seen": 783680, + "step": 2925 + }, + { + "epoch": 0.7680209698558322, + "grad_norm": 1.7513716220855713, + "learning_rate": 3.8387942332896464e-05, + "loss": 0.5015, + "num_input_tokens_seen": 784912, + "step": 2930 + }, + { + "epoch": 0.7693315858453473, + "grad_norm": 1.6991792917251587, + "learning_rate": 3.8453473132372215e-05, + "loss": 0.538, + "num_input_tokens_seen": 786368, + "step": 2935 + }, + { + "epoch": 0.7706422018348624, + "grad_norm": 3.2183473110198975, + "learning_rate": 3.851900393184797e-05, + "loss": 0.6837, + "num_input_tokens_seen": 787600, + "step": 2940 + }, + { + "epoch": 0.7719528178243774, + "grad_norm": 19.533300399780273, + "learning_rate": 3.858453473132372e-05, + "loss": 0.6916, + "num_input_tokens_seen": 788800, + "step": 2945 + }, + { + "epoch": 0.7732634338138925, + "grad_norm": 0.9509637951850891, + "learning_rate": 3.865006553079948e-05, + "loss": 0.6628, + "num_input_tokens_seen": 790464, + "step": 2950 + }, + { + "epoch": 0.7745740498034076, + "grad_norm": 2.3997325897216797, + "learning_rate": 3.871559633027523e-05, + "loss": 1.0472, + "num_input_tokens_seen": 791392, + "step": 2955 + }, + { + "epoch": 0.7758846657929227, + "grad_norm": 2.921187162399292, + "learning_rate": 3.878112712975098e-05, + "loss": 0.5231, + "num_input_tokens_seen": 792432, + "step": 2960 + }, + { + "epoch": 0.7771952817824378, + "grad_norm": 1.2485541105270386, + "learning_rate": 3.884665792922674e-05, + "loss": 0.556, + "num_input_tokens_seen": 794096, + "step": 2965 + }, + { + "epoch": 0.7785058977719528, + "grad_norm": 2.0704638957977295, + "learning_rate": 3.891218872870249e-05, + "loss": 0.4182, + "num_input_tokens_seen": 795664, + "step": 2970 + }, + { + "epoch": 0.7798165137614679, + "grad_norm": 2.503711700439453, + "learning_rate": 3.8977719528178246e-05, + "loss": 0.5619, + "num_input_tokens_seen": 796992, + "step": 2975 + }, + { + "epoch": 0.781127129750983, + "grad_norm": 1.1140791177749634, + "learning_rate": 3.9043250327654004e-05, + "loss": 0.3889, + "num_input_tokens_seen": 798144, + "step": 2980 + }, + { + "epoch": 0.7824377457404981, + "grad_norm": 1.526882529258728, + "learning_rate": 3.9108781127129754e-05, + "loss": 0.485, + "num_input_tokens_seen": 800224, + "step": 2985 + }, + { + "epoch": 0.783748361730013, + "grad_norm": 1.4799004793167114, + "learning_rate": 3.9174311926605505e-05, + "loss": 0.3899, + "num_input_tokens_seen": 801632, + "step": 2990 + }, + { + "epoch": 0.7850589777195282, + "grad_norm": 15.891334533691406, + "learning_rate": 3.923984272608126e-05, + "loss": 0.7386, + "num_input_tokens_seen": 803088, + "step": 2995 + }, + { + "epoch": 0.7863695937090432, + "grad_norm": 1.8680166006088257, + "learning_rate": 3.930537352555701e-05, + "loss": 0.645, + "num_input_tokens_seen": 804240, + "step": 3000 + }, + { + "epoch": 0.7876802096985583, + "grad_norm": 3.455017566680908, + "learning_rate": 3.937090432503276e-05, + "loss": 0.7149, + "num_input_tokens_seen": 805312, + "step": 3005 + }, + { + "epoch": 0.7889908256880734, + "grad_norm": 1.5347683429718018, + "learning_rate": 3.943643512450852e-05, + "loss": 0.2985, + "num_input_tokens_seen": 807184, + "step": 3010 + }, + { + "epoch": 0.7903014416775884, + "grad_norm": 2.0526678562164307, + "learning_rate": 3.950196592398428e-05, + "loss": 0.622, + "num_input_tokens_seen": 808400, + "step": 3015 + }, + { + "epoch": 0.7916120576671035, + "grad_norm": 2.139221429824829, + "learning_rate": 3.956749672346003e-05, + "loss": 0.5211, + "num_input_tokens_seen": 809936, + "step": 3020 + }, + { + "epoch": 0.7929226736566186, + "grad_norm": 1.1268168687820435, + "learning_rate": 3.9633027522935785e-05, + "loss": 0.5383, + "num_input_tokens_seen": 811216, + "step": 3025 + }, + { + "epoch": 0.7942332896461337, + "grad_norm": 1.4741157293319702, + "learning_rate": 3.9698558322411536e-05, + "loss": 0.4741, + "num_input_tokens_seen": 812288, + "step": 3030 + }, + { + "epoch": 0.7955439056356488, + "grad_norm": 3.3533430099487305, + "learning_rate": 3.9764089121887286e-05, + "loss": 0.5008, + "num_input_tokens_seen": 813536, + "step": 3035 + }, + { + "epoch": 0.7968545216251638, + "grad_norm": 3.2856509685516357, + "learning_rate": 3.9829619921363044e-05, + "loss": 0.5714, + "num_input_tokens_seen": 814768, + "step": 3040 + }, + { + "epoch": 0.7981651376146789, + "grad_norm": 1.1569148302078247, + "learning_rate": 3.9895150720838794e-05, + "loss": 0.3647, + "num_input_tokens_seen": 815984, + "step": 3045 + }, + { + "epoch": 0.799475753604194, + "grad_norm": 1.0658525228500366, + "learning_rate": 3.996068152031455e-05, + "loss": 0.4115, + "num_input_tokens_seen": 817744, + "step": 3050 + }, + { + "epoch": 0.8007863695937091, + "grad_norm": 1.8702068328857422, + "learning_rate": 4.00262123197903e-05, + "loss": 0.5042, + "num_input_tokens_seen": 818944, + "step": 3055 + }, + { + "epoch": 0.8020969855832241, + "grad_norm": 1.591485619544983, + "learning_rate": 4.009174311926606e-05, + "loss": 0.4527, + "num_input_tokens_seen": 820240, + "step": 3060 + }, + { + "epoch": 0.8034076015727392, + "grad_norm": 1.1479222774505615, + "learning_rate": 4.015727391874181e-05, + "loss": 0.4555, + "num_input_tokens_seen": 822032, + "step": 3065 + }, + { + "epoch": 0.8047182175622543, + "grad_norm": 1.9368969202041626, + "learning_rate": 4.022280471821757e-05, + "loss": 0.4286, + "num_input_tokens_seen": 823536, + "step": 3070 + }, + { + "epoch": 0.8060288335517694, + "grad_norm": 0.5284149646759033, + "learning_rate": 4.028833551769332e-05, + "loss": 0.5715, + "num_input_tokens_seen": 824864, + "step": 3075 + }, + { + "epoch": 0.8073394495412844, + "grad_norm": 1.7267143726348877, + "learning_rate": 4.035386631716907e-05, + "loss": 0.575, + "num_input_tokens_seen": 825936, + "step": 3080 + }, + { + "epoch": 0.8086500655307994, + "grad_norm": 0.853394091129303, + "learning_rate": 4.0419397116644826e-05, + "loss": 0.5968, + "num_input_tokens_seen": 826960, + "step": 3085 + }, + { + "epoch": 0.8099606815203145, + "grad_norm": 5.799693584442139, + "learning_rate": 4.048492791612058e-05, + "loss": 0.2536, + "num_input_tokens_seen": 827840, + "step": 3090 + }, + { + "epoch": 0.8112712975098296, + "grad_norm": 2.1639606952667236, + "learning_rate": 4.0550458715596333e-05, + "loss": 0.573, + "num_input_tokens_seen": 829056, + "step": 3095 + }, + { + "epoch": 0.8125819134993447, + "grad_norm": 0.9499208927154541, + "learning_rate": 4.0615989515072084e-05, + "loss": 0.339, + "num_input_tokens_seen": 830528, + "step": 3100 + }, + { + "epoch": 0.8138925294888598, + "grad_norm": 3.4559402465820312, + "learning_rate": 4.068152031454784e-05, + "loss": 0.5273, + "num_input_tokens_seen": 831920, + "step": 3105 + }, + { + "epoch": 0.8152031454783748, + "grad_norm": 2.6881537437438965, + "learning_rate": 4.074705111402359e-05, + "loss": 0.537, + "num_input_tokens_seen": 835104, + "step": 3110 + }, + { + "epoch": 0.8165137614678899, + "grad_norm": 1.1537561416625977, + "learning_rate": 4.081258191349935e-05, + "loss": 0.3954, + "num_input_tokens_seen": 837472, + "step": 3115 + }, + { + "epoch": 0.817824377457405, + "grad_norm": 1.773888349533081, + "learning_rate": 4.08781127129751e-05, + "loss": 0.4218, + "num_input_tokens_seen": 838688, + "step": 3120 + }, + { + "epoch": 0.8191349934469201, + "grad_norm": 1.6166276931762695, + "learning_rate": 4.094364351245085e-05, + "loss": 0.5962, + "num_input_tokens_seen": 840448, + "step": 3125 + }, + { + "epoch": 0.8204456094364351, + "grad_norm": 1.1005845069885254, + "learning_rate": 4.100917431192661e-05, + "loss": 0.3767, + "num_input_tokens_seen": 841712, + "step": 3130 + }, + { + "epoch": 0.8217562254259502, + "grad_norm": 2.338071346282959, + "learning_rate": 4.1074705111402365e-05, + "loss": 0.8721, + "num_input_tokens_seen": 843024, + "step": 3135 + }, + { + "epoch": 0.8230668414154653, + "grad_norm": 0.9270979762077332, + "learning_rate": 4.1140235910878115e-05, + "loss": 0.7216, + "num_input_tokens_seen": 844736, + "step": 3140 + }, + { + "epoch": 0.8243774574049804, + "grad_norm": 0.7281689047813416, + "learning_rate": 4.1205766710353866e-05, + "loss": 0.4774, + "num_input_tokens_seen": 846432, + "step": 3145 + }, + { + "epoch": 0.8256880733944955, + "grad_norm": 1.1513400077819824, + "learning_rate": 4.127129750982962e-05, + "loss": 0.5695, + "num_input_tokens_seen": 847984, + "step": 3150 + }, + { + "epoch": 0.8269986893840104, + "grad_norm": 1.359252691268921, + "learning_rate": 4.1336828309305374e-05, + "loss": 0.5457, + "num_input_tokens_seen": 849072, + "step": 3155 + }, + { + "epoch": 0.8283093053735255, + "grad_norm": 1.2369157075881958, + "learning_rate": 4.140235910878113e-05, + "loss": 0.5563, + "num_input_tokens_seen": 850496, + "step": 3160 + }, + { + "epoch": 0.8296199213630406, + "grad_norm": 63.382102966308594, + "learning_rate": 4.146788990825689e-05, + "loss": 0.3604, + "num_input_tokens_seen": 851904, + "step": 3165 + }, + { + "epoch": 0.8309305373525557, + "grad_norm": 3.1818950176239014, + "learning_rate": 4.153342070773263e-05, + "loss": 0.4989, + "num_input_tokens_seen": 852864, + "step": 3170 + }, + { + "epoch": 0.8322411533420708, + "grad_norm": 1.5154732465744019, + "learning_rate": 4.159895150720839e-05, + "loss": 0.5712, + "num_input_tokens_seen": 854032, + "step": 3175 + }, + { + "epoch": 0.8335517693315858, + "grad_norm": 1.891648292541504, + "learning_rate": 4.1664482306684147e-05, + "loss": 0.483, + "num_input_tokens_seen": 855120, + "step": 3180 + }, + { + "epoch": 0.8348623853211009, + "grad_norm": 0.6425706148147583, + "learning_rate": 4.17300131061599e-05, + "loss": 0.4685, + "num_input_tokens_seen": 856320, + "step": 3185 + }, + { + "epoch": 0.836173001310616, + "grad_norm": 1.0114758014678955, + "learning_rate": 4.1795543905635654e-05, + "loss": 0.3534, + "num_input_tokens_seen": 857520, + "step": 3190 + }, + { + "epoch": 0.8374836173001311, + "grad_norm": 0.9350383281707764, + "learning_rate": 4.1861074705111405e-05, + "loss": 0.3624, + "num_input_tokens_seen": 858864, + "step": 3195 + }, + { + "epoch": 0.8387942332896461, + "grad_norm": 1.2706356048583984, + "learning_rate": 4.1926605504587155e-05, + "loss": 0.5725, + "num_input_tokens_seen": 861424, + "step": 3200 + }, + { + "epoch": 0.8401048492791612, + "grad_norm": 2.1104576587677, + "learning_rate": 4.199213630406291e-05, + "loss": 0.8135, + "num_input_tokens_seen": 862448, + "step": 3205 + }, + { + "epoch": 0.8414154652686763, + "grad_norm": 0.744979202747345, + "learning_rate": 4.205766710353867e-05, + "loss": 0.4708, + "num_input_tokens_seen": 863408, + "step": 3210 + }, + { + "epoch": 0.8427260812581914, + "grad_norm": 2.344285726547241, + "learning_rate": 4.2123197903014414e-05, + "loss": 0.4995, + "num_input_tokens_seen": 864432, + "step": 3215 + }, + { + "epoch": 0.8440366972477065, + "grad_norm": 2.0049502849578857, + "learning_rate": 4.218872870249017e-05, + "loss": 0.403, + "num_input_tokens_seen": 865648, + "step": 3220 + }, + { + "epoch": 0.8453473132372215, + "grad_norm": 1.436206579208374, + "learning_rate": 4.225425950196593e-05, + "loss": 0.4282, + "num_input_tokens_seen": 867536, + "step": 3225 + }, + { + "epoch": 0.8466579292267365, + "grad_norm": 1.4365112781524658, + "learning_rate": 4.231979030144168e-05, + "loss": 0.5901, + "num_input_tokens_seen": 868864, + "step": 3230 + }, + { + "epoch": 0.8479685452162516, + "grad_norm": 1.1281468868255615, + "learning_rate": 4.2385321100917436e-05, + "loss": 0.6637, + "num_input_tokens_seen": 870272, + "step": 3235 + }, + { + "epoch": 0.8492791612057667, + "grad_norm": 7.130519390106201, + "learning_rate": 4.245085190039319e-05, + "loss": 0.6849, + "num_input_tokens_seen": 871344, + "step": 3240 + }, + { + "epoch": 0.8505897771952818, + "grad_norm": 2.947800397872925, + "learning_rate": 4.251638269986894e-05, + "loss": 0.5599, + "num_input_tokens_seen": 872848, + "step": 3245 + }, + { + "epoch": 0.8519003931847968, + "grad_norm": 2.0177319049835205, + "learning_rate": 4.2581913499344695e-05, + "loss": 0.6501, + "num_input_tokens_seen": 874160, + "step": 3250 + }, + { + "epoch": 0.8532110091743119, + "grad_norm": 1.0661530494689941, + "learning_rate": 4.264744429882045e-05, + "loss": 0.3642, + "num_input_tokens_seen": 875424, + "step": 3255 + }, + { + "epoch": 0.854521625163827, + "grad_norm": 3.635067939758301, + "learning_rate": 4.27129750982962e-05, + "loss": 0.63, + "num_input_tokens_seen": 876544, + "step": 3260 + }, + { + "epoch": 0.8558322411533421, + "grad_norm": 1.5630171298980713, + "learning_rate": 4.277850589777195e-05, + "loss": 0.4559, + "num_input_tokens_seen": 878032, + "step": 3265 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 1.1472007036209106, + "learning_rate": 4.284403669724771e-05, + "loss": 0.6221, + "num_input_tokens_seen": 880112, + "step": 3270 + }, + { + "epoch": 0.8584534731323722, + "grad_norm": 1.1535052061080933, + "learning_rate": 4.290956749672346e-05, + "loss": 0.4155, + "num_input_tokens_seen": 881520, + "step": 3275 + }, + { + "epoch": 0.8597640891218873, + "grad_norm": 1.3986231088638306, + "learning_rate": 4.297509829619922e-05, + "loss": 0.4087, + "num_input_tokens_seen": 882592, + "step": 3280 + }, + { + "epoch": 0.8610747051114024, + "grad_norm": 1.660356044769287, + "learning_rate": 4.304062909567497e-05, + "loss": 0.442, + "num_input_tokens_seen": 883792, + "step": 3285 + }, + { + "epoch": 0.8623853211009175, + "grad_norm": 1.6402145624160767, + "learning_rate": 4.310615989515072e-05, + "loss": 0.5626, + "num_input_tokens_seen": 885200, + "step": 3290 + }, + { + "epoch": 0.8636959370904325, + "grad_norm": 1.109967589378357, + "learning_rate": 4.3171690694626476e-05, + "loss": 0.4815, + "num_input_tokens_seen": 886752, + "step": 3295 + }, + { + "epoch": 0.8650065530799476, + "grad_norm": 2.1888692378997803, + "learning_rate": 4.3237221494102234e-05, + "loss": 0.576, + "num_input_tokens_seen": 888080, + "step": 3300 + }, + { + "epoch": 0.8663171690694627, + "grad_norm": 1.8846129179000854, + "learning_rate": 4.3302752293577984e-05, + "loss": 0.673, + "num_input_tokens_seen": 889248, + "step": 3305 + }, + { + "epoch": 0.8676277850589778, + "grad_norm": 1.558648705482483, + "learning_rate": 4.3368283093053735e-05, + "loss": 0.5583, + "num_input_tokens_seen": 890368, + "step": 3310 + }, + { + "epoch": 0.8689384010484927, + "grad_norm": 2.06126070022583, + "learning_rate": 4.343381389252949e-05, + "loss": 0.6351, + "num_input_tokens_seen": 891472, + "step": 3315 + }, + { + "epoch": 0.8702490170380078, + "grad_norm": 2.030460834503174, + "learning_rate": 4.349934469200524e-05, + "loss": 0.5092, + "num_input_tokens_seen": 892688, + "step": 3320 + }, + { + "epoch": 0.8715596330275229, + "grad_norm": 1.273772954940796, + "learning_rate": 4.3564875491481e-05, + "loss": 0.568, + "num_input_tokens_seen": 893920, + "step": 3325 + }, + { + "epoch": 0.872870249017038, + "grad_norm": 9.176977157592773, + "learning_rate": 4.363040629095675e-05, + "loss": 0.4485, + "num_input_tokens_seen": 895072, + "step": 3330 + }, + { + "epoch": 0.8741808650065531, + "grad_norm": 1.148756742477417, + "learning_rate": 4.36959370904325e-05, + "loss": 0.5706, + "num_input_tokens_seen": 896272, + "step": 3335 + }, + { + "epoch": 0.8754914809960681, + "grad_norm": 0.8094926476478577, + "learning_rate": 4.376146788990826e-05, + "loss": 0.6412, + "num_input_tokens_seen": 897648, + "step": 3340 + }, + { + "epoch": 0.8768020969855832, + "grad_norm": 2.0549542903900146, + "learning_rate": 4.3826998689384016e-05, + "loss": 0.7167, + "num_input_tokens_seen": 898768, + "step": 3345 + }, + { + "epoch": 0.8781127129750983, + "grad_norm": 0.9087781310081482, + "learning_rate": 4.3892529488859766e-05, + "loss": 0.5439, + "num_input_tokens_seen": 900240, + "step": 3350 + }, + { + "epoch": 0.8794233289646134, + "grad_norm": 0.6473121047019958, + "learning_rate": 4.395806028833552e-05, + "loss": 0.5357, + "num_input_tokens_seen": 902000, + "step": 3355 + }, + { + "epoch": 0.8807339449541285, + "grad_norm": 2.195338487625122, + "learning_rate": 4.4023591087811274e-05, + "loss": 0.5592, + "num_input_tokens_seen": 903488, + "step": 3360 + }, + { + "epoch": 0.8820445609436435, + "grad_norm": 0.9735397100448608, + "learning_rate": 4.4089121887287024e-05, + "loss": 0.4966, + "num_input_tokens_seen": 904736, + "step": 3365 + }, + { + "epoch": 0.8833551769331586, + "grad_norm": 1.1517996788024902, + "learning_rate": 4.415465268676278e-05, + "loss": 0.4486, + "num_input_tokens_seen": 907040, + "step": 3370 + }, + { + "epoch": 0.8846657929226737, + "grad_norm": 3.848196506500244, + "learning_rate": 4.422018348623854e-05, + "loss": 0.6384, + "num_input_tokens_seen": 908128, + "step": 3375 + }, + { + "epoch": 0.8859764089121888, + "grad_norm": 1.168552041053772, + "learning_rate": 4.428571428571428e-05, + "loss": 0.6205, + "num_input_tokens_seen": 909712, + "step": 3380 + }, + { + "epoch": 0.8872870249017037, + "grad_norm": 1.0526094436645508, + "learning_rate": 4.435124508519004e-05, + "loss": 0.6551, + "num_input_tokens_seen": 910944, + "step": 3385 + }, + { + "epoch": 0.8885976408912188, + "grad_norm": 1.0369672775268555, + "learning_rate": 4.44167758846658e-05, + "loss": 0.5018, + "num_input_tokens_seen": 912096, + "step": 3390 + }, + { + "epoch": 0.8899082568807339, + "grad_norm": 1.63604736328125, + "learning_rate": 4.448230668414155e-05, + "loss": 0.6398, + "num_input_tokens_seen": 913248, + "step": 3395 + }, + { + "epoch": 0.891218872870249, + "grad_norm": 3.122770309448242, + "learning_rate": 4.45478374836173e-05, + "loss": 0.4601, + "num_input_tokens_seen": 914288, + "step": 3400 + }, + { + "epoch": 0.8925294888597641, + "grad_norm": 1.4494860172271729, + "learning_rate": 4.4613368283093056e-05, + "loss": 0.5067, + "num_input_tokens_seen": 915312, + "step": 3405 + }, + { + "epoch": 0.8938401048492791, + "grad_norm": 1.0662459135055542, + "learning_rate": 4.4678899082568806e-05, + "loss": 0.3875, + "num_input_tokens_seen": 916624, + "step": 3410 + }, + { + "epoch": 0.8951507208387942, + "grad_norm": 1.6901447772979736, + "learning_rate": 4.4744429882044564e-05, + "loss": 0.478, + "num_input_tokens_seen": 918048, + "step": 3415 + }, + { + "epoch": 0.8964613368283093, + "grad_norm": 1.0041804313659668, + "learning_rate": 4.480996068152032e-05, + "loss": 0.7344, + "num_input_tokens_seen": 919472, + "step": 3420 + }, + { + "epoch": 0.8977719528178244, + "grad_norm": 2.956575870513916, + "learning_rate": 4.487549148099607e-05, + "loss": 0.5555, + "num_input_tokens_seen": 920688, + "step": 3425 + }, + { + "epoch": 0.8990825688073395, + "grad_norm": 3.026740312576294, + "learning_rate": 4.494102228047182e-05, + "loss": 0.6508, + "num_input_tokens_seen": 921456, + "step": 3430 + }, + { + "epoch": 0.9003931847968545, + "grad_norm": 1.4601582288742065, + "learning_rate": 4.500655307994758e-05, + "loss": 0.4621, + "num_input_tokens_seen": 922816, + "step": 3435 + }, + { + "epoch": 0.9017038007863696, + "grad_norm": 1.3332034349441528, + "learning_rate": 4.507208387942333e-05, + "loss": 0.8513, + "num_input_tokens_seen": 924304, + "step": 3440 + }, + { + "epoch": 0.9030144167758847, + "grad_norm": 2.2536134719848633, + "learning_rate": 4.513761467889909e-05, + "loss": 0.461, + "num_input_tokens_seen": 925696, + "step": 3445 + }, + { + "epoch": 0.9043250327653998, + "grad_norm": 1.9460529088974, + "learning_rate": 4.520314547837484e-05, + "loss": 0.7884, + "num_input_tokens_seen": 926688, + "step": 3450 + }, + { + "epoch": 0.9056356487549148, + "grad_norm": 1.6076067686080933, + "learning_rate": 4.526867627785059e-05, + "loss": 0.7816, + "num_input_tokens_seen": 928336, + "step": 3455 + }, + { + "epoch": 0.9069462647444299, + "grad_norm": 0.8514717221260071, + "learning_rate": 4.5334207077326345e-05, + "loss": 0.3352, + "num_input_tokens_seen": 929664, + "step": 3460 + }, + { + "epoch": 0.908256880733945, + "grad_norm": 2.3616364002227783, + "learning_rate": 4.53997378768021e-05, + "loss": 0.6602, + "num_input_tokens_seen": 931216, + "step": 3465 + }, + { + "epoch": 0.90956749672346, + "grad_norm": 88.99708557128906, + "learning_rate": 4.546526867627785e-05, + "loss": 1.1129, + "num_input_tokens_seen": 932512, + "step": 3470 + }, + { + "epoch": 0.9108781127129751, + "grad_norm": 18.81009864807129, + "learning_rate": 4.5530799475753604e-05, + "loss": 1.5403, + "num_input_tokens_seen": 933440, + "step": 3475 + }, + { + "epoch": 0.9121887287024901, + "grad_norm": 1.0865709781646729, + "learning_rate": 4.559633027522936e-05, + "loss": 0.5091, + "num_input_tokens_seen": 934736, + "step": 3480 + }, + { + "epoch": 0.9134993446920052, + "grad_norm": 1.423156976699829, + "learning_rate": 4.566186107470511e-05, + "loss": 0.3343, + "num_input_tokens_seen": 935920, + "step": 3485 + }, + { + "epoch": 0.9148099606815203, + "grad_norm": 2.298848867416382, + "learning_rate": 4.572739187418087e-05, + "loss": 0.593, + "num_input_tokens_seen": 937392, + "step": 3490 + }, + { + "epoch": 0.9161205766710354, + "grad_norm": 1.5283048152923584, + "learning_rate": 4.579292267365662e-05, + "loss": 0.4771, + "num_input_tokens_seen": 938560, + "step": 3495 + }, + { + "epoch": 0.9174311926605505, + "grad_norm": 1.843017339706421, + "learning_rate": 4.585845347313238e-05, + "loss": 0.6115, + "num_input_tokens_seen": 939776, + "step": 3500 + }, + { + "epoch": 0.9187418086500655, + "grad_norm": 5.772060871124268, + "learning_rate": 4.592398427260813e-05, + "loss": 0.7145, + "num_input_tokens_seen": 940960, + "step": 3505 + }, + { + "epoch": 0.9200524246395806, + "grad_norm": 0.7023653984069824, + "learning_rate": 4.5989515072083885e-05, + "loss": 0.4551, + "num_input_tokens_seen": 942384, + "step": 3510 + }, + { + "epoch": 0.9213630406290957, + "grad_norm": 1.9862502813339233, + "learning_rate": 4.6055045871559635e-05, + "loss": 0.5551, + "num_input_tokens_seen": 943872, + "step": 3515 + }, + { + "epoch": 0.9226736566186108, + "grad_norm": 2.4456543922424316, + "learning_rate": 4.6120576671035386e-05, + "loss": 0.6122, + "num_input_tokens_seen": 945360, + "step": 3520 + }, + { + "epoch": 0.9239842726081258, + "grad_norm": 3.5326645374298096, + "learning_rate": 4.618610747051114e-05, + "loss": 0.4408, + "num_input_tokens_seen": 946480, + "step": 3525 + }, + { + "epoch": 0.9252948885976409, + "grad_norm": 1.7323780059814453, + "learning_rate": 4.6251638269986893e-05, + "loss": 0.5234, + "num_input_tokens_seen": 947744, + "step": 3530 + }, + { + "epoch": 0.926605504587156, + "grad_norm": 1.5126473903656006, + "learning_rate": 4.631716906946265e-05, + "loss": 0.65, + "num_input_tokens_seen": 948976, + "step": 3535 + }, + { + "epoch": 0.927916120576671, + "grad_norm": 1.0435340404510498, + "learning_rate": 4.63826998689384e-05, + "loss": 0.5835, + "num_input_tokens_seen": 950384, + "step": 3540 + }, + { + "epoch": 0.9292267365661862, + "grad_norm": 1.3728597164154053, + "learning_rate": 4.644823066841416e-05, + "loss": 0.5243, + "num_input_tokens_seen": 951584, + "step": 3545 + }, + { + "epoch": 0.9305373525557011, + "grad_norm": 1.2229400873184204, + "learning_rate": 4.651376146788991e-05, + "loss": 0.6499, + "num_input_tokens_seen": 952912, + "step": 3550 + }, + { + "epoch": 0.9318479685452162, + "grad_norm": 1.2310445308685303, + "learning_rate": 4.6579292267365666e-05, + "loss": 0.7328, + "num_input_tokens_seen": 954368, + "step": 3555 + }, + { + "epoch": 0.9331585845347313, + "grad_norm": 0.9580561518669128, + "learning_rate": 4.664482306684142e-05, + "loss": 0.4747, + "num_input_tokens_seen": 956000, + "step": 3560 + }, + { + "epoch": 0.9344692005242464, + "grad_norm": 1.1881502866744995, + "learning_rate": 4.671035386631717e-05, + "loss": 0.5145, + "num_input_tokens_seen": 957328, + "step": 3565 + }, + { + "epoch": 0.9357798165137615, + "grad_norm": 1.2048046588897705, + "learning_rate": 4.6775884665792925e-05, + "loss": 0.5319, + "num_input_tokens_seen": 958528, + "step": 3570 + }, + { + "epoch": 0.9370904325032765, + "grad_norm": 6.79465389251709, + "learning_rate": 4.684141546526868e-05, + "loss": 0.9308, + "num_input_tokens_seen": 959824, + "step": 3575 + }, + { + "epoch": 0.9384010484927916, + "grad_norm": 1.0672916173934937, + "learning_rate": 4.690694626474443e-05, + "loss": 0.4791, + "num_input_tokens_seen": 961232, + "step": 3580 + }, + { + "epoch": 0.9397116644823067, + "grad_norm": 1.33601713180542, + "learning_rate": 4.697247706422019e-05, + "loss": 0.6643, + "num_input_tokens_seen": 962480, + "step": 3585 + }, + { + "epoch": 0.9410222804718218, + "grad_norm": 0.6035807132720947, + "learning_rate": 4.703800786369594e-05, + "loss": 0.2748, + "num_input_tokens_seen": 964128, + "step": 3590 + }, + { + "epoch": 0.9423328964613368, + "grad_norm": 2.2531418800354004, + "learning_rate": 4.710353866317169e-05, + "loss": 0.6014, + "num_input_tokens_seen": 965584, + "step": 3595 + }, + { + "epoch": 0.9436435124508519, + "grad_norm": 2.404212474822998, + "learning_rate": 4.716906946264745e-05, + "loss": 0.5212, + "num_input_tokens_seen": 967136, + "step": 3600 + }, + { + "epoch": 0.944954128440367, + "grad_norm": 1.86550772190094, + "learning_rate": 4.72346002621232e-05, + "loss": 0.6029, + "num_input_tokens_seen": 968512, + "step": 3605 + }, + { + "epoch": 0.9462647444298821, + "grad_norm": 1.9536539316177368, + "learning_rate": 4.730013106159895e-05, + "loss": 0.7484, + "num_input_tokens_seen": 969744, + "step": 3610 + }, + { + "epoch": 0.9475753604193972, + "grad_norm": 1.0291041135787964, + "learning_rate": 4.7365661861074707e-05, + "loss": 0.3888, + "num_input_tokens_seen": 970976, + "step": 3615 + }, + { + "epoch": 0.9488859764089121, + "grad_norm": 1.7485733032226562, + "learning_rate": 4.7431192660550464e-05, + "loss": 0.4555, + "num_input_tokens_seen": 972240, + "step": 3620 + }, + { + "epoch": 0.9501965923984272, + "grad_norm": 2.483531951904297, + "learning_rate": 4.7496723460026214e-05, + "loss": 0.5738, + "num_input_tokens_seen": 973904, + "step": 3625 + }, + { + "epoch": 0.9515072083879423, + "grad_norm": 0.981083333492279, + "learning_rate": 4.756225425950197e-05, + "loss": 0.4021, + "num_input_tokens_seen": 975392, + "step": 3630 + }, + { + "epoch": 0.9528178243774574, + "grad_norm": 2.5220611095428467, + "learning_rate": 4.762778505897772e-05, + "loss": 0.4658, + "num_input_tokens_seen": 976560, + "step": 3635 + }, + { + "epoch": 0.9541284403669725, + "grad_norm": 1.7529041767120361, + "learning_rate": 4.769331585845347e-05, + "loss": 0.7576, + "num_input_tokens_seen": 977600, + "step": 3640 + }, + { + "epoch": 0.9554390563564875, + "grad_norm": 4.517643928527832, + "learning_rate": 4.775884665792923e-05, + "loss": 0.6324, + "num_input_tokens_seen": 978896, + "step": 3645 + }, + { + "epoch": 0.9567496723460026, + "grad_norm": 1.3473387956619263, + "learning_rate": 4.782437745740499e-05, + "loss": 0.4969, + "num_input_tokens_seen": 980320, + "step": 3650 + }, + { + "epoch": 0.9580602883355177, + "grad_norm": 1.3736492395401, + "learning_rate": 4.788990825688074e-05, + "loss": 0.7315, + "num_input_tokens_seen": 981744, + "step": 3655 + }, + { + "epoch": 0.9593709043250328, + "grad_norm": 1.688493251800537, + "learning_rate": 4.795543905635649e-05, + "loss": 0.4462, + "num_input_tokens_seen": 983088, + "step": 3660 + }, + { + "epoch": 0.9606815203145478, + "grad_norm": 0.900262176990509, + "learning_rate": 4.8020969855832246e-05, + "loss": 0.4159, + "num_input_tokens_seen": 984320, + "step": 3665 + }, + { + "epoch": 0.9619921363040629, + "grad_norm": 0.8219708204269409, + "learning_rate": 4.8086500655307996e-05, + "loss": 0.3933, + "num_input_tokens_seen": 985520, + "step": 3670 + }, + { + "epoch": 0.963302752293578, + "grad_norm": 1.0377558469772339, + "learning_rate": 4.8152031454783754e-05, + "loss": 0.4487, + "num_input_tokens_seen": 986464, + "step": 3675 + }, + { + "epoch": 0.9646133682830931, + "grad_norm": 0.8458212614059448, + "learning_rate": 4.8217562254259504e-05, + "loss": 0.3602, + "num_input_tokens_seen": 987680, + "step": 3680 + }, + { + "epoch": 0.9659239842726082, + "grad_norm": 1.4225391149520874, + "learning_rate": 4.8283093053735255e-05, + "loss": 0.4343, + "num_input_tokens_seen": 989408, + "step": 3685 + }, + { + "epoch": 0.9672346002621232, + "grad_norm": 3.759155750274658, + "learning_rate": 4.834862385321101e-05, + "loss": 0.3545, + "num_input_tokens_seen": 990368, + "step": 3690 + }, + { + "epoch": 0.9685452162516383, + "grad_norm": 1.507443904876709, + "learning_rate": 4.841415465268677e-05, + "loss": 0.4737, + "num_input_tokens_seen": 992112, + "step": 3695 + }, + { + "epoch": 0.9698558322411533, + "grad_norm": 1.862964153289795, + "learning_rate": 4.847968545216252e-05, + "loss": 0.5261, + "num_input_tokens_seen": 993488, + "step": 3700 + }, + { + "epoch": 0.9711664482306684, + "grad_norm": 1.521420955657959, + "learning_rate": 4.854521625163827e-05, + "loss": 0.5846, + "num_input_tokens_seen": 994800, + "step": 3705 + }, + { + "epoch": 0.9724770642201835, + "grad_norm": 0.849865734577179, + "learning_rate": 4.861074705111403e-05, + "loss": 0.4845, + "num_input_tokens_seen": 996192, + "step": 3710 + }, + { + "epoch": 0.9737876802096985, + "grad_norm": 1.4495189189910889, + "learning_rate": 4.867627785058978e-05, + "loss": 0.426, + "num_input_tokens_seen": 997872, + "step": 3715 + }, + { + "epoch": 0.9750982961992136, + "grad_norm": 2.4648685455322266, + "learning_rate": 4.8741808650065535e-05, + "loss": 0.6116, + "num_input_tokens_seen": 999520, + "step": 3720 + }, + { + "epoch": 0.9764089121887287, + "grad_norm": 1.1273245811462402, + "learning_rate": 4.880733944954129e-05, + "loss": 0.5332, + "num_input_tokens_seen": 1000496, + "step": 3725 + }, + { + "epoch": 0.9777195281782438, + "grad_norm": 0.5055431723594666, + "learning_rate": 4.8872870249017036e-05, + "loss": 0.4836, + "num_input_tokens_seen": 1001472, + "step": 3730 + }, + { + "epoch": 0.9790301441677588, + "grad_norm": 1.343143105506897, + "learning_rate": 4.8938401048492794e-05, + "loss": 0.5067, + "num_input_tokens_seen": 1002688, + "step": 3735 + }, + { + "epoch": 0.9803407601572739, + "grad_norm": 1.5341612100601196, + "learning_rate": 4.900393184796855e-05, + "loss": 0.5911, + "num_input_tokens_seen": 1004000, + "step": 3740 + }, + { + "epoch": 0.981651376146789, + "grad_norm": 3.962782859802246, + "learning_rate": 4.90694626474443e-05, + "loss": 0.5997, + "num_input_tokens_seen": 1005744, + "step": 3745 + }, + { + "epoch": 0.9829619921363041, + "grad_norm": 0.8794181942939758, + "learning_rate": 4.913499344692005e-05, + "loss": 0.5351, + "num_input_tokens_seen": 1007344, + "step": 3750 + }, + { + "epoch": 0.9842726081258192, + "grad_norm": 2.3766520023345947, + "learning_rate": 4.920052424639581e-05, + "loss": 0.3958, + "num_input_tokens_seen": 1008736, + "step": 3755 + }, + { + "epoch": 0.9855832241153342, + "grad_norm": 1.6082688570022583, + "learning_rate": 4.926605504587156e-05, + "loss": 0.3917, + "num_input_tokens_seen": 1009808, + "step": 3760 + }, + { + "epoch": 0.9868938401048493, + "grad_norm": 1.146477460861206, + "learning_rate": 4.933158584534732e-05, + "loss": 0.4175, + "num_input_tokens_seen": 1010688, + "step": 3765 + }, + { + "epoch": 0.9882044560943644, + "grad_norm": 1.4731441736221313, + "learning_rate": 4.9397116644823075e-05, + "loss": 0.5057, + "num_input_tokens_seen": 1012320, + "step": 3770 + }, + { + "epoch": 0.9895150720838795, + "grad_norm": 1.347402811050415, + "learning_rate": 4.946264744429882e-05, + "loss": 0.3895, + "num_input_tokens_seen": 1013328, + "step": 3775 + }, + { + "epoch": 0.9908256880733946, + "grad_norm": 2.113576650619507, + "learning_rate": 4.9528178243774576e-05, + "loss": 0.4538, + "num_input_tokens_seen": 1014912, + "step": 3780 + }, + { + "epoch": 0.9921363040629095, + "grad_norm": 1.726593255996704, + "learning_rate": 4.959370904325033e-05, + "loss": 0.4071, + "num_input_tokens_seen": 1016336, + "step": 3785 + }, + { + "epoch": 0.9934469200524246, + "grad_norm": 0.49665576219558716, + "learning_rate": 4.9659239842726083e-05, + "loss": 0.6299, + "num_input_tokens_seen": 1017568, + "step": 3790 + }, + { + "epoch": 0.9947575360419397, + "grad_norm": 0.9893030524253845, + "learning_rate": 4.972477064220184e-05, + "loss": 0.3962, + "num_input_tokens_seen": 1018736, + "step": 3795 + }, + { + "epoch": 0.9960681520314548, + "grad_norm": 3.766287088394165, + "learning_rate": 4.979030144167759e-05, + "loss": 0.3867, + "num_input_tokens_seen": 1019744, + "step": 3800 + }, + { + "epoch": 0.9973787680209698, + "grad_norm": 1.3951705694198608, + "learning_rate": 4.985583224115334e-05, + "loss": 0.6028, + "num_input_tokens_seen": 1020880, + "step": 3805 + }, + { + "epoch": 0.9986893840104849, + "grad_norm": 1.2803809642791748, + "learning_rate": 4.99213630406291e-05, + "loss": 0.6953, + "num_input_tokens_seen": 1022400, + "step": 3810 + }, + { + "epoch": 1.0, + "grad_norm": 3.01516056060791, + "learning_rate": 4.9986893840104856e-05, + "loss": 0.5661, + "num_input_tokens_seen": 1023568, + "step": 3815 + }, + { + "epoch": 1.000262123197903, + "eval_loss": 0.6061103343963623, + "eval_runtime": 16.8848, + "eval_samples_per_second": 50.223, + "eval_steps_per_second": 25.111, + "num_input_tokens_seen": 1023840, + "step": 3816 + }, + { + "epoch": 1.001310615989515, + "grad_norm": 7.159420490264893, + "learning_rate": 4.9999998325613814e-05, + "loss": 0.4425, + "num_input_tokens_seen": 1024656, + "step": 3820 + }, + { + "epoch": 1.0026212319790302, + "grad_norm": 1.0801142454147339, + "learning_rate": 4.999999152342031e-05, + "loss": 0.5058, + "num_input_tokens_seen": 1025904, + "step": 3825 + }, + { + "epoch": 1.0039318479685453, + "grad_norm": 1.3498189449310303, + "learning_rate": 4.999997948877178e-05, + "loss": 0.4553, + "num_input_tokens_seen": 1026816, + "step": 3830 + }, + { + "epoch": 1.0052424639580604, + "grad_norm": 1.378947377204895, + "learning_rate": 4.9999962221670737e-05, + "loss": 0.5867, + "num_input_tokens_seen": 1028176, + "step": 3835 + }, + { + "epoch": 1.0065530799475753, + "grad_norm": 8.686419486999512, + "learning_rate": 4.9999939722120804e-05, + "loss": 0.4607, + "num_input_tokens_seen": 1029232, + "step": 3840 + }, + { + "epoch": 1.0078636959370904, + "grad_norm": 3.099161148071289, + "learning_rate": 4.9999911990126674e-05, + "loss": 0.4068, + "num_input_tokens_seen": 1030512, + "step": 3845 + }, + { + "epoch": 1.0091743119266054, + "grad_norm": 1.4740298986434937, + "learning_rate": 4.999987902569416e-05, + "loss": 0.3592, + "num_input_tokens_seen": 1031488, + "step": 3850 + }, + { + "epoch": 1.0104849279161205, + "grad_norm": 0.8074439167976379, + "learning_rate": 4.9999840828830167e-05, + "loss": 0.3932, + "num_input_tokens_seen": 1033152, + "step": 3855 + }, + { + "epoch": 1.0117955439056356, + "grad_norm": 1.1634896993637085, + "learning_rate": 4.999979739954269e-05, + "loss": 0.4552, + "num_input_tokens_seen": 1034624, + "step": 3860 + }, + { + "epoch": 1.0131061598951507, + "grad_norm": 0.7311257123947144, + "learning_rate": 4.999974873784081e-05, + "loss": 0.4133, + "num_input_tokens_seen": 1035584, + "step": 3865 + }, + { + "epoch": 1.0144167758846658, + "grad_norm": 1.243216633796692, + "learning_rate": 4.9999694843734715e-05, + "loss": 0.477, + "num_input_tokens_seen": 1036960, + "step": 3870 + }, + { + "epoch": 1.015727391874181, + "grad_norm": 0.6374621987342834, + "learning_rate": 4.999963571723568e-05, + "loss": 0.5576, + "num_input_tokens_seen": 1038720, + "step": 3875 + }, + { + "epoch": 1.017038007863696, + "grad_norm": 0.5909344553947449, + "learning_rate": 4.9999571358356095e-05, + "loss": 0.4132, + "num_input_tokens_seen": 1040400, + "step": 3880 + }, + { + "epoch": 1.018348623853211, + "grad_norm": 0.6123512983322144, + "learning_rate": 4.9999501767109416e-05, + "loss": 0.3842, + "num_input_tokens_seen": 1042304, + "step": 3885 + }, + { + "epoch": 1.019659239842726, + "grad_norm": 1.0722328424453735, + "learning_rate": 4.999942694351021e-05, + "loss": 0.571, + "num_input_tokens_seen": 1043728, + "step": 3890 + }, + { + "epoch": 1.020969855832241, + "grad_norm": 1.1478902101516724, + "learning_rate": 4.9999346887574146e-05, + "loss": 0.4474, + "num_input_tokens_seen": 1044752, + "step": 3895 + }, + { + "epoch": 1.0222804718217562, + "grad_norm": 1.1567803621292114, + "learning_rate": 4.999926159931797e-05, + "loss": 0.5764, + "num_input_tokens_seen": 1045968, + "step": 3900 + }, + { + "epoch": 1.0235910878112713, + "grad_norm": 1.372678279876709, + "learning_rate": 4.999917107875954e-05, + "loss": 0.5223, + "num_input_tokens_seen": 1046848, + "step": 3905 + }, + { + "epoch": 1.0249017038007864, + "grad_norm": 2.2322804927825928, + "learning_rate": 4.999907532591779e-05, + "loss": 0.5639, + "num_input_tokens_seen": 1048208, + "step": 3910 + }, + { + "epoch": 1.0262123197903015, + "grad_norm": 1.9682461023330688, + "learning_rate": 4.9998974340812775e-05, + "loss": 0.3614, + "num_input_tokens_seen": 1049744, + "step": 3915 + }, + { + "epoch": 1.0275229357798166, + "grad_norm": 1.7464532852172852, + "learning_rate": 4.999886812346563e-05, + "loss": 0.5128, + "num_input_tokens_seen": 1050816, + "step": 3920 + }, + { + "epoch": 1.0288335517693317, + "grad_norm": 2.5073115825653076, + "learning_rate": 4.999875667389858e-05, + "loss": 0.4542, + "num_input_tokens_seen": 1051888, + "step": 3925 + }, + { + "epoch": 1.0301441677588468, + "grad_norm": 1.9288241863250732, + "learning_rate": 4.999863999213495e-05, + "loss": 0.2777, + "num_input_tokens_seen": 1052960, + "step": 3930 + }, + { + "epoch": 1.0314547837483616, + "grad_norm": 1.6619274616241455, + "learning_rate": 4.999851807819916e-05, + "loss": 0.4618, + "num_input_tokens_seen": 1054592, + "step": 3935 + }, + { + "epoch": 1.0327653997378767, + "grad_norm": 0.7285422086715698, + "learning_rate": 4.9998390932116734e-05, + "loss": 0.4172, + "num_input_tokens_seen": 1056080, + "step": 3940 + }, + { + "epoch": 1.0340760157273918, + "grad_norm": 1.3092368841171265, + "learning_rate": 4.9998258553914286e-05, + "loss": 0.464, + "num_input_tokens_seen": 1057520, + "step": 3945 + }, + { + "epoch": 1.035386631716907, + "grad_norm": 0.853740930557251, + "learning_rate": 4.999812094361951e-05, + "loss": 0.6422, + "num_input_tokens_seen": 1058928, + "step": 3950 + }, + { + "epoch": 1.036697247706422, + "grad_norm": 0.9385638236999512, + "learning_rate": 4.9997978101261215e-05, + "loss": 0.2569, + "num_input_tokens_seen": 1060400, + "step": 3955 + }, + { + "epoch": 1.0380078636959371, + "grad_norm": 1.094931721687317, + "learning_rate": 4.9997830026869294e-05, + "loss": 0.5309, + "num_input_tokens_seen": 1061824, + "step": 3960 + }, + { + "epoch": 1.0393184796854522, + "grad_norm": 0.9077247381210327, + "learning_rate": 4.999767672047474e-05, + "loss": 0.7101, + "num_input_tokens_seen": 1063584, + "step": 3965 + }, + { + "epoch": 1.0406290956749673, + "grad_norm": 1.9578781127929688, + "learning_rate": 4.999751818210965e-05, + "loss": 0.5739, + "num_input_tokens_seen": 1064736, + "step": 3970 + }, + { + "epoch": 1.0419397116644824, + "grad_norm": 1.3156464099884033, + "learning_rate": 4.999735441180718e-05, + "loss": 0.501, + "num_input_tokens_seen": 1065744, + "step": 3975 + }, + { + "epoch": 1.0432503276539973, + "grad_norm": 2.0846142768859863, + "learning_rate": 4.999718540960164e-05, + "loss": 0.5147, + "num_input_tokens_seen": 1068432, + "step": 3980 + }, + { + "epoch": 1.0445609436435124, + "grad_norm": 7.6723737716674805, + "learning_rate": 4.999701117552839e-05, + "loss": 0.5059, + "num_input_tokens_seen": 1069712, + "step": 3985 + }, + { + "epoch": 1.0458715596330275, + "grad_norm": 2.3094489574432373, + "learning_rate": 4.999683170962388e-05, + "loss": 0.3664, + "num_input_tokens_seen": 1071328, + "step": 3990 + }, + { + "epoch": 1.0471821756225426, + "grad_norm": 2.2947518825531006, + "learning_rate": 4.9996647011925685e-05, + "loss": 0.7255, + "num_input_tokens_seen": 1072288, + "step": 3995 + }, + { + "epoch": 1.0484927916120577, + "grad_norm": 1.2076008319854736, + "learning_rate": 4.9996457082472465e-05, + "loss": 0.8107, + "num_input_tokens_seen": 1073872, + "step": 4000 + }, + { + "epoch": 1.0498034076015728, + "grad_norm": 2.4826090335845947, + "learning_rate": 4.999626192130397e-05, + "loss": 0.3473, + "num_input_tokens_seen": 1075728, + "step": 4005 + }, + { + "epoch": 1.0511140235910879, + "grad_norm": 0.7189033031463623, + "learning_rate": 4.999606152846104e-05, + "loss": 0.3985, + "num_input_tokens_seen": 1076864, + "step": 4010 + }, + { + "epoch": 1.052424639580603, + "grad_norm": 2.4521825313568115, + "learning_rate": 4.9995855903985624e-05, + "loss": 0.3926, + "num_input_tokens_seen": 1078192, + "step": 4015 + }, + { + "epoch": 1.053735255570118, + "grad_norm": 1.4424529075622559, + "learning_rate": 4.9995645047920756e-05, + "loss": 0.5633, + "num_input_tokens_seen": 1079280, + "step": 4020 + }, + { + "epoch": 1.0550458715596331, + "grad_norm": 1.178566575050354, + "learning_rate": 4.999542896031056e-05, + "loss": 0.4701, + "num_input_tokens_seen": 1080672, + "step": 4025 + }, + { + "epoch": 1.056356487549148, + "grad_norm": 1.259565830230713, + "learning_rate": 4.999520764120029e-05, + "loss": 0.4689, + "num_input_tokens_seen": 1082208, + "step": 4030 + }, + { + "epoch": 1.0576671035386631, + "grad_norm": 2.3262879848480225, + "learning_rate": 4.999498109063624e-05, + "loss": 0.4146, + "num_input_tokens_seen": 1083584, + "step": 4035 + }, + { + "epoch": 1.0589777195281782, + "grad_norm": 1.5530219078063965, + "learning_rate": 4.999474930866583e-05, + "loss": 0.4572, + "num_input_tokens_seen": 1084864, + "step": 4040 + }, + { + "epoch": 1.0602883355176933, + "grad_norm": 1.105785846710205, + "learning_rate": 4.999451229533759e-05, + "loss": 0.5478, + "num_input_tokens_seen": 1085872, + "step": 4045 + }, + { + "epoch": 1.0615989515072084, + "grad_norm": 0.6894363760948181, + "learning_rate": 4.99942700507011e-05, + "loss": 0.5742, + "num_input_tokens_seen": 1087408, + "step": 4050 + }, + { + "epoch": 1.0629095674967235, + "grad_norm": 1.3715085983276367, + "learning_rate": 4.9994022574807085e-05, + "loss": 0.5452, + "num_input_tokens_seen": 1088576, + "step": 4055 + }, + { + "epoch": 1.0642201834862386, + "grad_norm": 0.8880573511123657, + "learning_rate": 4.999376986770733e-05, + "loss": 0.3743, + "num_input_tokens_seen": 1090144, + "step": 4060 + }, + { + "epoch": 1.0655307994757537, + "grad_norm": 6.831776142120361, + "learning_rate": 4.999351192945473e-05, + "loss": 0.6008, + "num_input_tokens_seen": 1091296, + "step": 4065 + }, + { + "epoch": 1.0668414154652686, + "grad_norm": 1.0674132108688354, + "learning_rate": 4.999324876010326e-05, + "loss": 0.3977, + "num_input_tokens_seen": 1092448, + "step": 4070 + }, + { + "epoch": 1.0681520314547837, + "grad_norm": 0.5760433673858643, + "learning_rate": 4.999298035970801e-05, + "loss": 0.3594, + "num_input_tokens_seen": 1093552, + "step": 4075 + }, + { + "epoch": 1.0694626474442988, + "grad_norm": 0.7781333923339844, + "learning_rate": 4.999270672832516e-05, + "loss": 0.4645, + "num_input_tokens_seen": 1094640, + "step": 4080 + }, + { + "epoch": 1.0707732634338138, + "grad_norm": 2.2601654529571533, + "learning_rate": 4.9992427866011974e-05, + "loss": 0.556, + "num_input_tokens_seen": 1096288, + "step": 4085 + }, + { + "epoch": 1.072083879423329, + "grad_norm": 0.6371235847473145, + "learning_rate": 4.9992143772826826e-05, + "loss": 0.4866, + "num_input_tokens_seen": 1097888, + "step": 4090 + }, + { + "epoch": 1.073394495412844, + "grad_norm": 4.473040580749512, + "learning_rate": 4.9991854448829165e-05, + "loss": 0.5002, + "num_input_tokens_seen": 1099248, + "step": 4095 + }, + { + "epoch": 1.0747051114023591, + "grad_norm": 0.437052845954895, + "learning_rate": 4.999155989407954e-05, + "loss": 0.7568, + "num_input_tokens_seen": 1100736, + "step": 4100 + }, + { + "epoch": 1.0760157273918742, + "grad_norm": 1.888143539428711, + "learning_rate": 4.999126010863963e-05, + "loss": 0.4022, + "num_input_tokens_seen": 1102208, + "step": 4105 + }, + { + "epoch": 1.0773263433813893, + "grad_norm": 1.5007684230804443, + "learning_rate": 4.999095509257214e-05, + "loss": 0.4156, + "num_input_tokens_seen": 1103216, + "step": 4110 + }, + { + "epoch": 1.0786369593709044, + "grad_norm": 0.9805920720100403, + "learning_rate": 4.999064484594095e-05, + "loss": 0.3341, + "num_input_tokens_seen": 1104400, + "step": 4115 + }, + { + "epoch": 1.0799475753604193, + "grad_norm": 1.7725881338119507, + "learning_rate": 4.999032936881096e-05, + "loss": 0.7361, + "num_input_tokens_seen": 1105552, + "step": 4120 + }, + { + "epoch": 1.0812581913499344, + "grad_norm": 3.1542410850524902, + "learning_rate": 4.999000866124822e-05, + "loss": 0.7484, + "num_input_tokens_seen": 1107616, + "step": 4125 + }, + { + "epoch": 1.0825688073394495, + "grad_norm": 1.1477241516113281, + "learning_rate": 4.9989682723319846e-05, + "loss": 0.5231, + "num_input_tokens_seen": 1109056, + "step": 4130 + }, + { + "epoch": 1.0838794233289646, + "grad_norm": 0.6805821061134338, + "learning_rate": 4.9989351555094055e-05, + "loss": 0.4521, + "num_input_tokens_seen": 1110400, + "step": 4135 + }, + { + "epoch": 1.0851900393184797, + "grad_norm": 1.7783633470535278, + "learning_rate": 4.998901515664017e-05, + "loss": 0.4684, + "num_input_tokens_seen": 1111536, + "step": 4140 + }, + { + "epoch": 1.0865006553079948, + "grad_norm": 0.6347336769104004, + "learning_rate": 4.998867352802858e-05, + "loss": 0.4447, + "num_input_tokens_seen": 1113936, + "step": 4145 + }, + { + "epoch": 1.0878112712975099, + "grad_norm": 8.215877532958984, + "learning_rate": 4.9988326669330797e-05, + "loss": 0.4758, + "num_input_tokens_seen": 1115312, + "step": 4150 + }, + { + "epoch": 1.089121887287025, + "grad_norm": 1.4193700551986694, + "learning_rate": 4.998797458061942e-05, + "loss": 0.4573, + "num_input_tokens_seen": 1116384, + "step": 4155 + }, + { + "epoch": 1.09043250327654, + "grad_norm": 0.7294561266899109, + "learning_rate": 4.998761726196814e-05, + "loss": 0.6298, + "num_input_tokens_seen": 1117360, + "step": 4160 + }, + { + "epoch": 1.091743119266055, + "grad_norm": 1.4981597661972046, + "learning_rate": 4.9987254713451754e-05, + "loss": 0.393, + "num_input_tokens_seen": 1118544, + "step": 4165 + }, + { + "epoch": 1.09305373525557, + "grad_norm": 1.242059350013733, + "learning_rate": 4.998688693514612e-05, + "loss": 0.585, + "num_input_tokens_seen": 1119824, + "step": 4170 + }, + { + "epoch": 1.0943643512450851, + "grad_norm": 0.5649529099464417, + "learning_rate": 4.998651392712823e-05, + "loss": 0.5521, + "num_input_tokens_seen": 1121264, + "step": 4175 + }, + { + "epoch": 1.0956749672346002, + "grad_norm": 1.285164713859558, + "learning_rate": 4.998613568947614e-05, + "loss": 0.3416, + "num_input_tokens_seen": 1122560, + "step": 4180 + }, + { + "epoch": 1.0969855832241153, + "grad_norm": 2.247164726257324, + "learning_rate": 4.998575222226903e-05, + "loss": 0.477, + "num_input_tokens_seen": 1123920, + "step": 4185 + }, + { + "epoch": 1.0982961992136304, + "grad_norm": 0.6770831346511841, + "learning_rate": 4.9985363525587145e-05, + "loss": 0.3895, + "num_input_tokens_seen": 1125280, + "step": 4190 + }, + { + "epoch": 1.0996068152031455, + "grad_norm": 1.232072353363037, + "learning_rate": 4.998496959951185e-05, + "loss": 0.3661, + "num_input_tokens_seen": 1127056, + "step": 4195 + }, + { + "epoch": 1.1009174311926606, + "grad_norm": 1.538723349571228, + "learning_rate": 4.998457044412559e-05, + "loss": 0.6114, + "num_input_tokens_seen": 1128720, + "step": 4200 + }, + { + "epoch": 1.1022280471821757, + "grad_norm": 0.633808434009552, + "learning_rate": 4.998416605951191e-05, + "loss": 0.3396, + "num_input_tokens_seen": 1129968, + "step": 4205 + }, + { + "epoch": 1.1035386631716908, + "grad_norm": 0.9805379509925842, + "learning_rate": 4.998375644575543e-05, + "loss": 0.4634, + "num_input_tokens_seen": 1131408, + "step": 4210 + }, + { + "epoch": 1.1048492791612057, + "grad_norm": 0.9327640533447266, + "learning_rate": 4.99833416029419e-05, + "loss": 0.4183, + "num_input_tokens_seen": 1132400, + "step": 4215 + }, + { + "epoch": 1.1061598951507208, + "grad_norm": 0.992981493473053, + "learning_rate": 4.998292153115814e-05, + "loss": 0.5609, + "num_input_tokens_seen": 1133712, + "step": 4220 + }, + { + "epoch": 1.1074705111402359, + "grad_norm": 0.7763444185256958, + "learning_rate": 4.9982496230492085e-05, + "loss": 0.5013, + "num_input_tokens_seen": 1135232, + "step": 4225 + }, + { + "epoch": 1.108781127129751, + "grad_norm": 1.3405827283859253, + "learning_rate": 4.998206570103271e-05, + "loss": 0.6096, + "num_input_tokens_seen": 1136288, + "step": 4230 + }, + { + "epoch": 1.110091743119266, + "grad_norm": 1.6410553455352783, + "learning_rate": 4.998162994287017e-05, + "loss": 0.4467, + "num_input_tokens_seen": 1137584, + "step": 4235 + }, + { + "epoch": 1.1114023591087812, + "grad_norm": 0.6157506704330444, + "learning_rate": 4.998118895609564e-05, + "loss": 0.4461, + "num_input_tokens_seen": 1139408, + "step": 4240 + }, + { + "epoch": 1.1127129750982963, + "grad_norm": 6.133310317993164, + "learning_rate": 4.9980742740801425e-05, + "loss": 0.6515, + "num_input_tokens_seen": 1140448, + "step": 4245 + }, + { + "epoch": 1.1140235910878113, + "grad_norm": 1.2706414461135864, + "learning_rate": 4.998029129708092e-05, + "loss": 0.4232, + "num_input_tokens_seen": 1142000, + "step": 4250 + }, + { + "epoch": 1.1153342070773264, + "grad_norm": 0.4760555922985077, + "learning_rate": 4.997983462502861e-05, + "loss": 0.4713, + "num_input_tokens_seen": 1143568, + "step": 4255 + }, + { + "epoch": 1.1166448230668413, + "grad_norm": 1.7915475368499756, + "learning_rate": 4.997937272474007e-05, + "loss": 0.3882, + "num_input_tokens_seen": 1144800, + "step": 4260 + }, + { + "epoch": 1.1179554390563564, + "grad_norm": 1.6390833854675293, + "learning_rate": 4.9978905596311985e-05, + "loss": 0.3961, + "num_input_tokens_seen": 1146080, + "step": 4265 + }, + { + "epoch": 1.1192660550458715, + "grad_norm": 0.7380208373069763, + "learning_rate": 4.9978433239842115e-05, + "loss": 0.4202, + "num_input_tokens_seen": 1147488, + "step": 4270 + }, + { + "epoch": 1.1205766710353866, + "grad_norm": 1.2072010040283203, + "learning_rate": 4.997795565542933e-05, + "loss": 0.8581, + "num_input_tokens_seen": 1148368, + "step": 4275 + }, + { + "epoch": 1.1218872870249017, + "grad_norm": 0.7859166264533997, + "learning_rate": 4.997747284317358e-05, + "loss": 0.5372, + "num_input_tokens_seen": 1149808, + "step": 4280 + }, + { + "epoch": 1.1231979030144168, + "grad_norm": 6.115128993988037, + "learning_rate": 4.997698480317593e-05, + "loss": 0.587, + "num_input_tokens_seen": 1151184, + "step": 4285 + }, + { + "epoch": 1.124508519003932, + "grad_norm": 4.987656593322754, + "learning_rate": 4.997649153553851e-05, + "loss": 0.528, + "num_input_tokens_seen": 1152288, + "step": 4290 + }, + { + "epoch": 1.125819134993447, + "grad_norm": 1.5860340595245361, + "learning_rate": 4.9975993040364574e-05, + "loss": 0.5426, + "num_input_tokens_seen": 1153648, + "step": 4295 + }, + { + "epoch": 1.127129750982962, + "grad_norm": 0.6857008337974548, + "learning_rate": 4.9975489317758446e-05, + "loss": 0.6093, + "num_input_tokens_seen": 1154864, + "step": 4300 + }, + { + "epoch": 1.1284403669724772, + "grad_norm": 1.3244496583938599, + "learning_rate": 4.9974980367825555e-05, + "loss": 0.5505, + "num_input_tokens_seen": 1156608, + "step": 4305 + }, + { + "epoch": 1.129750982961992, + "grad_norm": 2.280834674835205, + "learning_rate": 4.997446619067243e-05, + "loss": 0.4126, + "num_input_tokens_seen": 1157968, + "step": 4310 + }, + { + "epoch": 1.1310615989515072, + "grad_norm": 4.0228352546691895, + "learning_rate": 4.997394678640669e-05, + "loss": 0.7822, + "num_input_tokens_seen": 1159152, + "step": 4315 + }, + { + "epoch": 1.1323722149410222, + "grad_norm": 1.034698724746704, + "learning_rate": 4.997342215513703e-05, + "loss": 0.5023, + "num_input_tokens_seen": 1160144, + "step": 4320 + }, + { + "epoch": 1.1336828309305373, + "grad_norm": 1.382405161857605, + "learning_rate": 4.9972892296973263e-05, + "loss": 0.4647, + "num_input_tokens_seen": 1161136, + "step": 4325 + }, + { + "epoch": 1.1349934469200524, + "grad_norm": 1.6055073738098145, + "learning_rate": 4.997235721202629e-05, + "loss": 0.5846, + "num_input_tokens_seen": 1162384, + "step": 4330 + }, + { + "epoch": 1.1363040629095675, + "grad_norm": 0.8906358480453491, + "learning_rate": 4.997181690040811e-05, + "loss": 0.4929, + "num_input_tokens_seen": 1163696, + "step": 4335 + }, + { + "epoch": 1.1376146788990826, + "grad_norm": 3.538158416748047, + "learning_rate": 4.997127136223179e-05, + "loss": 0.5408, + "num_input_tokens_seen": 1164864, + "step": 4340 + }, + { + "epoch": 1.1389252948885977, + "grad_norm": 1.2898240089416504, + "learning_rate": 4.997072059761153e-05, + "loss": 0.382, + "num_input_tokens_seen": 1165968, + "step": 4345 + }, + { + "epoch": 1.1402359108781126, + "grad_norm": 1.4341918230056763, + "learning_rate": 4.997016460666258e-05, + "loss": 0.5557, + "num_input_tokens_seen": 1167488, + "step": 4350 + }, + { + "epoch": 1.1415465268676277, + "grad_norm": 0.6775271892547607, + "learning_rate": 4.996960338950134e-05, + "loss": 0.7966, + "num_input_tokens_seen": 1169024, + "step": 4355 + }, + { + "epoch": 1.1428571428571428, + "grad_norm": 3.671318292617798, + "learning_rate": 4.996903694624525e-05, + "loss": 0.3896, + "num_input_tokens_seen": 1170352, + "step": 4360 + }, + { + "epoch": 1.144167758846658, + "grad_norm": 1.013917088508606, + "learning_rate": 4.996846527701287e-05, + "loss": 0.574, + "num_input_tokens_seen": 1171808, + "step": 4365 + }, + { + "epoch": 1.145478374836173, + "grad_norm": 1.137479305267334, + "learning_rate": 4.996788838192384e-05, + "loss": 0.5173, + "num_input_tokens_seen": 1173104, + "step": 4370 + }, + { + "epoch": 1.146788990825688, + "grad_norm": 4.104167461395264, + "learning_rate": 4.9967306261098924e-05, + "loss": 0.5372, + "num_input_tokens_seen": 1174208, + "step": 4375 + }, + { + "epoch": 1.1480996068152032, + "grad_norm": 22.062149047851562, + "learning_rate": 4.996671891465994e-05, + "loss": 0.8105, + "num_input_tokens_seen": 1176528, + "step": 4380 + }, + { + "epoch": 1.1494102228047183, + "grad_norm": 2.0132479667663574, + "learning_rate": 4.996612634272983e-05, + "loss": 0.885, + "num_input_tokens_seen": 1177936, + "step": 4385 + }, + { + "epoch": 1.1507208387942334, + "grad_norm": 1.173055648803711, + "learning_rate": 4.996552854543262e-05, + "loss": 0.4095, + "num_input_tokens_seen": 1178896, + "step": 4390 + }, + { + "epoch": 1.1520314547837485, + "grad_norm": 1.6593494415283203, + "learning_rate": 4.9964925522893416e-05, + "loss": 0.6971, + "num_input_tokens_seen": 1180160, + "step": 4395 + }, + { + "epoch": 1.1533420707732633, + "grad_norm": 1.1119199991226196, + "learning_rate": 4.9964317275238435e-05, + "loss": 0.3507, + "num_input_tokens_seen": 1181648, + "step": 4400 + }, + { + "epoch": 1.1546526867627784, + "grad_norm": 0.9521048665046692, + "learning_rate": 4.996370380259499e-05, + "loss": 0.9921, + "num_input_tokens_seen": 1182896, + "step": 4405 + }, + { + "epoch": 1.1559633027522935, + "grad_norm": 1.2328035831451416, + "learning_rate": 4.996308510509147e-05, + "loss": 0.6759, + "num_input_tokens_seen": 1184112, + "step": 4410 + }, + { + "epoch": 1.1572739187418086, + "grad_norm": 1.040117859840393, + "learning_rate": 4.9962461182857366e-05, + "loss": 0.3336, + "num_input_tokens_seen": 1185312, + "step": 4415 + }, + { + "epoch": 1.1585845347313237, + "grad_norm": 4.800933837890625, + "learning_rate": 4.9961832036023275e-05, + "loss": 0.4073, + "num_input_tokens_seen": 1186336, + "step": 4420 + }, + { + "epoch": 1.1598951507208388, + "grad_norm": 0.7837898135185242, + "learning_rate": 4.996119766472087e-05, + "loss": 0.6473, + "num_input_tokens_seen": 1187776, + "step": 4425 + }, + { + "epoch": 1.161205766710354, + "grad_norm": 2.193521022796631, + "learning_rate": 4.996055806908292e-05, + "loss": 0.4608, + "num_input_tokens_seen": 1189040, + "step": 4430 + }, + { + "epoch": 1.162516382699869, + "grad_norm": 1.5678430795669556, + "learning_rate": 4.99599132492433e-05, + "loss": 0.4416, + "num_input_tokens_seen": 1190272, + "step": 4435 + }, + { + "epoch": 1.163826998689384, + "grad_norm": 0.8347896933555603, + "learning_rate": 4.995926320533695e-05, + "loss": 0.5059, + "num_input_tokens_seen": 1191472, + "step": 4440 + }, + { + "epoch": 1.165137614678899, + "grad_norm": 0.5890889763832092, + "learning_rate": 4.995860793749995e-05, + "loss": 0.4633, + "num_input_tokens_seen": 1192832, + "step": 4445 + }, + { + "epoch": 1.166448230668414, + "grad_norm": 0.5925289988517761, + "learning_rate": 4.995794744586942e-05, + "loss": 0.5792, + "num_input_tokens_seen": 1194288, + "step": 4450 + }, + { + "epoch": 1.1677588466579292, + "grad_norm": 0.5788330435752869, + "learning_rate": 4.9957281730583635e-05, + "loss": 0.52, + "num_input_tokens_seen": 1195392, + "step": 4455 + }, + { + "epoch": 1.1690694626474443, + "grad_norm": 1.3121991157531738, + "learning_rate": 4.9956610791781885e-05, + "loss": 0.489, + "num_input_tokens_seen": 1196880, + "step": 4460 + }, + { + "epoch": 1.1703800786369594, + "grad_norm": 1.9262194633483887, + "learning_rate": 4.995593462960464e-05, + "loss": 0.5309, + "num_input_tokens_seen": 1198064, + "step": 4465 + }, + { + "epoch": 1.1716906946264745, + "grad_norm": 1.0025668144226074, + "learning_rate": 4.9955253244193375e-05, + "loss": 0.5672, + "num_input_tokens_seen": 1199152, + "step": 4470 + }, + { + "epoch": 1.1730013106159896, + "grad_norm": 1.0772182941436768, + "learning_rate": 4.9954566635690744e-05, + "loss": 0.5102, + "num_input_tokens_seen": 1200256, + "step": 4475 + }, + { + "epoch": 1.1743119266055047, + "grad_norm": 0.7799650430679321, + "learning_rate": 4.995387480424043e-05, + "loss": 0.5985, + "num_input_tokens_seen": 1201936, + "step": 4480 + }, + { + "epoch": 1.1756225425950197, + "grad_norm": 0.851098895072937, + "learning_rate": 4.995317774998723e-05, + "loss": 0.4225, + "num_input_tokens_seen": 1205088, + "step": 4485 + }, + { + "epoch": 1.1769331585845348, + "grad_norm": 1.13025963306427, + "learning_rate": 4.995247547307704e-05, + "loss": 0.4093, + "num_input_tokens_seen": 1206176, + "step": 4490 + }, + { + "epoch": 1.1782437745740497, + "grad_norm": 0.32524561882019043, + "learning_rate": 4.995176797365686e-05, + "loss": 0.4584, + "num_input_tokens_seen": 1207568, + "step": 4495 + }, + { + "epoch": 1.1795543905635648, + "grad_norm": 2.0631885528564453, + "learning_rate": 4.995105525187475e-05, + "loss": 0.3384, + "num_input_tokens_seen": 1208816, + "step": 4500 + }, + { + "epoch": 1.18086500655308, + "grad_norm": 1.2372745275497437, + "learning_rate": 4.9950337307879884e-05, + "loss": 0.6038, + "num_input_tokens_seen": 1209952, + "step": 4505 + }, + { + "epoch": 1.182175622542595, + "grad_norm": 0.6110579967498779, + "learning_rate": 4.994961414182254e-05, + "loss": 0.3846, + "num_input_tokens_seen": 1211408, + "step": 4510 + }, + { + "epoch": 1.18348623853211, + "grad_norm": 1.4522206783294678, + "learning_rate": 4.9948885753854056e-05, + "loss": 0.3498, + "num_input_tokens_seen": 1212816, + "step": 4515 + }, + { + "epoch": 1.1847968545216252, + "grad_norm": 1.0838282108306885, + "learning_rate": 4.99481521441269e-05, + "loss": 0.5078, + "num_input_tokens_seen": 1214048, + "step": 4520 + }, + { + "epoch": 1.1861074705111403, + "grad_norm": 1.1340490579605103, + "learning_rate": 4.99474133127946e-05, + "loss": 0.5216, + "num_input_tokens_seen": 1215360, + "step": 4525 + }, + { + "epoch": 1.1874180865006554, + "grad_norm": 1.0045467615127563, + "learning_rate": 4.99466692600118e-05, + "loss": 0.8024, + "num_input_tokens_seen": 1216352, + "step": 4530 + }, + { + "epoch": 1.1887287024901703, + "grad_norm": 0.6467545628547668, + "learning_rate": 4.994591998593423e-05, + "loss": 0.4992, + "num_input_tokens_seen": 1217680, + "step": 4535 + }, + { + "epoch": 1.1900393184796854, + "grad_norm": 1.7509610652923584, + "learning_rate": 4.9945165490718714e-05, + "loss": 0.4291, + "num_input_tokens_seen": 1219472, + "step": 4540 + }, + { + "epoch": 1.1913499344692005, + "grad_norm": 2.803439140319824, + "learning_rate": 4.994440577452316e-05, + "loss": 0.4501, + "num_input_tokens_seen": 1221008, + "step": 4545 + }, + { + "epoch": 1.1926605504587156, + "grad_norm": 2.8191802501678467, + "learning_rate": 4.9943640837506574e-05, + "loss": 0.6931, + "num_input_tokens_seen": 1222016, + "step": 4550 + }, + { + "epoch": 1.1939711664482306, + "grad_norm": 0.9519912600517273, + "learning_rate": 4.994287067982907e-05, + "loss": 0.5521, + "num_input_tokens_seen": 1223136, + "step": 4555 + }, + { + "epoch": 1.1952817824377457, + "grad_norm": 1.7243708372116089, + "learning_rate": 4.994209530165183e-05, + "loss": 0.9464, + "num_input_tokens_seen": 1224368, + "step": 4560 + }, + { + "epoch": 1.1965923984272608, + "grad_norm": 1.127439022064209, + "learning_rate": 4.9941314703137134e-05, + "loss": 0.4571, + "num_input_tokens_seen": 1226448, + "step": 4565 + }, + { + "epoch": 1.197903014416776, + "grad_norm": 0.7290454506874084, + "learning_rate": 4.994052888444837e-05, + "loss": 0.4697, + "num_input_tokens_seen": 1227792, + "step": 4570 + }, + { + "epoch": 1.199213630406291, + "grad_norm": 6.012959003448486, + "learning_rate": 4.993973784575e-05, + "loss": 0.4008, + "num_input_tokens_seen": 1229056, + "step": 4575 + }, + { + "epoch": 1.2005242463958061, + "grad_norm": 1.4813951253890991, + "learning_rate": 4.9938941587207586e-05, + "loss": 0.5003, + "num_input_tokens_seen": 1230272, + "step": 4580 + }, + { + "epoch": 1.2018348623853212, + "grad_norm": 0.7843748331069946, + "learning_rate": 4.9938140108987795e-05, + "loss": 0.7135, + "num_input_tokens_seen": 1231648, + "step": 4585 + }, + { + "epoch": 1.203145478374836, + "grad_norm": 1.025805115699768, + "learning_rate": 4.9937333411258366e-05, + "loss": 0.4789, + "num_input_tokens_seen": 1232736, + "step": 4590 + }, + { + "epoch": 1.2044560943643512, + "grad_norm": 1.084288239479065, + "learning_rate": 4.993652149418815e-05, + "loss": 0.3929, + "num_input_tokens_seen": 1233904, + "step": 4595 + }, + { + "epoch": 1.2057667103538663, + "grad_norm": 0.798294723033905, + "learning_rate": 4.9935704357947065e-05, + "loss": 0.6018, + "num_input_tokens_seen": 1235232, + "step": 4600 + }, + { + "epoch": 1.2070773263433814, + "grad_norm": 1.1851216554641724, + "learning_rate": 4.9934882002706154e-05, + "loss": 0.4629, + "num_input_tokens_seen": 1236608, + "step": 4605 + }, + { + "epoch": 1.2083879423328965, + "grad_norm": 0.7222753763198853, + "learning_rate": 4.9934054428637514e-05, + "loss": 0.6699, + "num_input_tokens_seen": 1237712, + "step": 4610 + }, + { + "epoch": 1.2096985583224116, + "grad_norm": 1.3623902797698975, + "learning_rate": 4.9933221635914365e-05, + "loss": 0.5151, + "num_input_tokens_seen": 1239152, + "step": 4615 + }, + { + "epoch": 1.2110091743119267, + "grad_norm": 1.0958704948425293, + "learning_rate": 4.9932383624711014e-05, + "loss": 0.336, + "num_input_tokens_seen": 1240784, + "step": 4620 + }, + { + "epoch": 1.2123197903014418, + "grad_norm": 4.012871742248535, + "learning_rate": 4.993154039520285e-05, + "loss": 0.6048, + "num_input_tokens_seen": 1242752, + "step": 4625 + }, + { + "epoch": 1.2136304062909566, + "grad_norm": 1.1000087261199951, + "learning_rate": 4.9930691947566354e-05, + "loss": 0.392, + "num_input_tokens_seen": 1244176, + "step": 4630 + }, + { + "epoch": 1.2149410222804717, + "grad_norm": 1.184188961982727, + "learning_rate": 4.992983828197911e-05, + "loss": 0.3537, + "num_input_tokens_seen": 1245376, + "step": 4635 + }, + { + "epoch": 1.2162516382699868, + "grad_norm": 1.3817336559295654, + "learning_rate": 4.99289793986198e-05, + "loss": 0.4621, + "num_input_tokens_seen": 1246544, + "step": 4640 + }, + { + "epoch": 1.217562254259502, + "grad_norm": 0.7936738133430481, + "learning_rate": 4.992811529766816e-05, + "loss": 0.4486, + "num_input_tokens_seen": 1248016, + "step": 4645 + }, + { + "epoch": 1.218872870249017, + "grad_norm": 1.9447717666625977, + "learning_rate": 4.992724597930508e-05, + "loss": 0.5156, + "num_input_tokens_seen": 1249392, + "step": 4650 + }, + { + "epoch": 1.2201834862385321, + "grad_norm": 0.6731107234954834, + "learning_rate": 4.9926371443712474e-05, + "loss": 0.462, + "num_input_tokens_seen": 1250960, + "step": 4655 + }, + { + "epoch": 1.2214941022280472, + "grad_norm": 3.585216522216797, + "learning_rate": 4.9925491691073396e-05, + "loss": 0.436, + "num_input_tokens_seen": 1252320, + "step": 4660 + }, + { + "epoch": 1.2228047182175623, + "grad_norm": 1.1433603763580322, + "learning_rate": 4.9924606721571975e-05, + "loss": 0.4713, + "num_input_tokens_seen": 1254080, + "step": 4665 + }, + { + "epoch": 1.2241153342070774, + "grad_norm": 0.524417519569397, + "learning_rate": 4.9923716535393434e-05, + "loss": 0.6017, + "num_input_tokens_seen": 1255168, + "step": 4670 + }, + { + "epoch": 1.2254259501965925, + "grad_norm": 0.8938450813293457, + "learning_rate": 4.9922821132724086e-05, + "loss": 0.5386, + "num_input_tokens_seen": 1256528, + "step": 4675 + }, + { + "epoch": 1.2267365661861074, + "grad_norm": 1.2157790660858154, + "learning_rate": 4.992192051375134e-05, + "loss": 0.4831, + "num_input_tokens_seen": 1257808, + "step": 4680 + }, + { + "epoch": 1.2280471821756225, + "grad_norm": 2.7477433681488037, + "learning_rate": 4.992101467866369e-05, + "loss": 0.5671, + "num_input_tokens_seen": 1259040, + "step": 4685 + }, + { + "epoch": 1.2293577981651376, + "grad_norm": 1.4498347043991089, + "learning_rate": 4.9920103627650734e-05, + "loss": 0.5357, + "num_input_tokens_seen": 1260592, + "step": 4690 + }, + { + "epoch": 1.2306684141546527, + "grad_norm": 1.2719151973724365, + "learning_rate": 4.991918736090314e-05, + "loss": 0.6257, + "num_input_tokens_seen": 1261680, + "step": 4695 + }, + { + "epoch": 1.2319790301441678, + "grad_norm": 3.442701578140259, + "learning_rate": 4.991826587861269e-05, + "loss": 0.6308, + "num_input_tokens_seen": 1262816, + "step": 4700 + }, + { + "epoch": 1.2332896461336829, + "grad_norm": 0.6952130794525146, + "learning_rate": 4.991733918097225e-05, + "loss": 0.5734, + "num_input_tokens_seen": 1264464, + "step": 4705 + }, + { + "epoch": 1.234600262123198, + "grad_norm": 1.9255977869033813, + "learning_rate": 4.9916407268175775e-05, + "loss": 0.4635, + "num_input_tokens_seen": 1265536, + "step": 4710 + }, + { + "epoch": 1.235910878112713, + "grad_norm": 3.3812594413757324, + "learning_rate": 4.991547014041831e-05, + "loss": 0.4616, + "num_input_tokens_seen": 1267104, + "step": 4715 + }, + { + "epoch": 1.237221494102228, + "grad_norm": 1.2395552396774292, + "learning_rate": 4.9914527797895995e-05, + "loss": 0.4191, + "num_input_tokens_seen": 1268256, + "step": 4720 + }, + { + "epoch": 1.238532110091743, + "grad_norm": 3.1497244834899902, + "learning_rate": 4.9913580240806054e-05, + "loss": 0.6099, + "num_input_tokens_seen": 1269184, + "step": 4725 + }, + { + "epoch": 1.2398427260812581, + "grad_norm": 2.9330270290374756, + "learning_rate": 4.991262746934682e-05, + "loss": 0.4836, + "num_input_tokens_seen": 1270768, + "step": 4730 + }, + { + "epoch": 1.2411533420707732, + "grad_norm": 1.2804139852523804, + "learning_rate": 4.991166948371771e-05, + "loss": 0.4186, + "num_input_tokens_seen": 1272336, + "step": 4735 + }, + { + "epoch": 1.2424639580602883, + "grad_norm": 0.8518379330635071, + "learning_rate": 4.991070628411921e-05, + "loss": 0.3702, + "num_input_tokens_seen": 1273520, + "step": 4740 + }, + { + "epoch": 1.2437745740498034, + "grad_norm": 1.5466947555541992, + "learning_rate": 4.9909737870752935e-05, + "loss": 0.3889, + "num_input_tokens_seen": 1274976, + "step": 4745 + }, + { + "epoch": 1.2450851900393185, + "grad_norm": 0.9309436082839966, + "learning_rate": 4.990876424382156e-05, + "loss": 0.524, + "num_input_tokens_seen": 1276096, + "step": 4750 + }, + { + "epoch": 1.2463958060288336, + "grad_norm": 1.1667503118515015, + "learning_rate": 4.9907785403528863e-05, + "loss": 0.3179, + "num_input_tokens_seen": 1278192, + "step": 4755 + }, + { + "epoch": 1.2477064220183487, + "grad_norm": 0.5746421813964844, + "learning_rate": 4.9906801350079724e-05, + "loss": 0.4082, + "num_input_tokens_seen": 1279520, + "step": 4760 + }, + { + "epoch": 1.2490170380078638, + "grad_norm": 2.0252106189727783, + "learning_rate": 4.9905812083680105e-05, + "loss": 0.5633, + "num_input_tokens_seen": 1280608, + "step": 4765 + }, + { + "epoch": 1.2503276539973789, + "grad_norm": 2.240793466567993, + "learning_rate": 4.990481760453704e-05, + "loss": 0.4318, + "num_input_tokens_seen": 1282416, + "step": 4770 + }, + { + "epoch": 1.2516382699868938, + "grad_norm": 2.1798205375671387, + "learning_rate": 4.9903817912858684e-05, + "loss": 0.6223, + "num_input_tokens_seen": 1283456, + "step": 4775 + }, + { + "epoch": 1.2529488859764089, + "grad_norm": 1.1037418842315674, + "learning_rate": 4.9902813008854274e-05, + "loss": 0.54, + "num_input_tokens_seen": 1284624, + "step": 4780 + }, + { + "epoch": 1.254259501965924, + "grad_norm": 1.2669802904129028, + "learning_rate": 4.990180289273413e-05, + "loss": 0.5168, + "num_input_tokens_seen": 1285888, + "step": 4785 + }, + { + "epoch": 1.255570117955439, + "grad_norm": 1.2606669664382935, + "learning_rate": 4.990078756470966e-05, + "loss": 0.3981, + "num_input_tokens_seen": 1287184, + "step": 4790 + }, + { + "epoch": 1.2568807339449541, + "grad_norm": 1.54890775680542, + "learning_rate": 4.989976702499339e-05, + "loss": 0.498, + "num_input_tokens_seen": 1288592, + "step": 4795 + }, + { + "epoch": 1.2581913499344692, + "grad_norm": 1.2783254384994507, + "learning_rate": 4.98987412737989e-05, + "loss": 0.7359, + "num_input_tokens_seen": 1289744, + "step": 4800 + }, + { + "epoch": 1.2595019659239843, + "grad_norm": 1.58345365524292, + "learning_rate": 4.989771031134088e-05, + "loss": 0.5692, + "num_input_tokens_seen": 1290960, + "step": 4805 + }, + { + "epoch": 1.2608125819134992, + "grad_norm": 1.5577095746994019, + "learning_rate": 4.989667413783511e-05, + "loss": 0.4865, + "num_input_tokens_seen": 1292208, + "step": 4810 + }, + { + "epoch": 1.2621231979030143, + "grad_norm": 0.6579248309135437, + "learning_rate": 4.9895632753498465e-05, + "loss": 0.5459, + "num_input_tokens_seen": 1293472, + "step": 4815 + }, + { + "epoch": 1.2634338138925294, + "grad_norm": 1.105115532875061, + "learning_rate": 4.98945861585489e-05, + "loss": 0.3869, + "num_input_tokens_seen": 1294640, + "step": 4820 + }, + { + "epoch": 1.2647444298820445, + "grad_norm": 2.097775936126709, + "learning_rate": 4.989353435320547e-05, + "loss": 0.5083, + "num_input_tokens_seen": 1296096, + "step": 4825 + }, + { + "epoch": 1.2660550458715596, + "grad_norm": 2.6927437782287598, + "learning_rate": 4.9892477337688306e-05, + "loss": 0.544, + "num_input_tokens_seen": 1297200, + "step": 4830 + }, + { + "epoch": 1.2673656618610747, + "grad_norm": 0.6970965266227722, + "learning_rate": 4.9891415112218654e-05, + "loss": 0.3522, + "num_input_tokens_seen": 1298512, + "step": 4835 + }, + { + "epoch": 1.2686762778505898, + "grad_norm": 1.5365101099014282, + "learning_rate": 4.9890347677018826e-05, + "loss": 0.5005, + "num_input_tokens_seen": 1299616, + "step": 4840 + }, + { + "epoch": 1.2699868938401049, + "grad_norm": 0.6475441455841064, + "learning_rate": 4.9889275032312235e-05, + "loss": 0.4123, + "num_input_tokens_seen": 1300784, + "step": 4845 + }, + { + "epoch": 1.27129750982962, + "grad_norm": 0.40982818603515625, + "learning_rate": 4.988819717832338e-05, + "loss": 0.4747, + "num_input_tokens_seen": 1302304, + "step": 4850 + }, + { + "epoch": 1.272608125819135, + "grad_norm": 0.9834890365600586, + "learning_rate": 4.988711411527786e-05, + "loss": 0.4463, + "num_input_tokens_seen": 1304224, + "step": 4855 + }, + { + "epoch": 1.2739187418086502, + "grad_norm": 0.6550593972206116, + "learning_rate": 4.988602584340236e-05, + "loss": 0.5948, + "num_input_tokens_seen": 1305584, + "step": 4860 + }, + { + "epoch": 1.2752293577981653, + "grad_norm": 0.9114547371864319, + "learning_rate": 4.988493236292465e-05, + "loss": 0.4847, + "num_input_tokens_seen": 1306848, + "step": 4865 + }, + { + "epoch": 1.2765399737876801, + "grad_norm": 1.3335719108581543, + "learning_rate": 4.98838336740736e-05, + "loss": 0.5456, + "num_input_tokens_seen": 1308064, + "step": 4870 + }, + { + "epoch": 1.2778505897771952, + "grad_norm": 0.70005202293396, + "learning_rate": 4.9882729777079154e-05, + "loss": 0.3834, + "num_input_tokens_seen": 1309152, + "step": 4875 + }, + { + "epoch": 1.2791612057667103, + "grad_norm": 1.4634032249450684, + "learning_rate": 4.9881620672172355e-05, + "loss": 0.502, + "num_input_tokens_seen": 1310352, + "step": 4880 + }, + { + "epoch": 1.2804718217562254, + "grad_norm": 0.8440139293670654, + "learning_rate": 4.988050635958535e-05, + "loss": 0.4293, + "num_input_tokens_seen": 1311408, + "step": 4885 + }, + { + "epoch": 1.2817824377457405, + "grad_norm": 1.1580003499984741, + "learning_rate": 4.987938683955135e-05, + "loss": 0.4994, + "num_input_tokens_seen": 1312656, + "step": 4890 + }, + { + "epoch": 1.2830930537352556, + "grad_norm": 1.7953540086746216, + "learning_rate": 4.987826211230467e-05, + "loss": 0.5005, + "num_input_tokens_seen": 1314096, + "step": 4895 + }, + { + "epoch": 1.2844036697247707, + "grad_norm": 0.71741783618927, + "learning_rate": 4.987713217808072e-05, + "loss": 0.3143, + "num_input_tokens_seen": 1315168, + "step": 4900 + }, + { + "epoch": 1.2857142857142856, + "grad_norm": 0.5136866569519043, + "learning_rate": 4.987599703711599e-05, + "loss": 0.6377, + "num_input_tokens_seen": 1316592, + "step": 4905 + }, + { + "epoch": 1.2870249017038007, + "grad_norm": 3.185080051422119, + "learning_rate": 4.9874856689648065e-05, + "loss": 0.5113, + "num_input_tokens_seen": 1317488, + "step": 4910 + }, + { + "epoch": 1.2883355176933158, + "grad_norm": 1.6140574216842651, + "learning_rate": 4.9873711135915605e-05, + "loss": 0.324, + "num_input_tokens_seen": 1319216, + "step": 4915 + }, + { + "epoch": 1.2896461336828309, + "grad_norm": 1.2355962991714478, + "learning_rate": 4.987256037615839e-05, + "loss": 0.5463, + "num_input_tokens_seen": 1320320, + "step": 4920 + }, + { + "epoch": 1.290956749672346, + "grad_norm": 0.9681946635246277, + "learning_rate": 4.987140441061726e-05, + "loss": 0.3903, + "num_input_tokens_seen": 1321664, + "step": 4925 + }, + { + "epoch": 1.292267365661861, + "grad_norm": 1.2405976057052612, + "learning_rate": 4.987024323953417e-05, + "loss": 0.32, + "num_input_tokens_seen": 1323040, + "step": 4930 + }, + { + "epoch": 1.2935779816513762, + "grad_norm": 0.7637469172477722, + "learning_rate": 4.9869076863152134e-05, + "loss": 0.3558, + "num_input_tokens_seen": 1324224, + "step": 4935 + }, + { + "epoch": 1.2948885976408913, + "grad_norm": 0.7355055809020996, + "learning_rate": 4.986790528171529e-05, + "loss": 0.3138, + "num_input_tokens_seen": 1325744, + "step": 4940 + }, + { + "epoch": 1.2961992136304064, + "grad_norm": 0.6957045197486877, + "learning_rate": 4.986672849546883e-05, + "loss": 0.4604, + "num_input_tokens_seen": 1327136, + "step": 4945 + }, + { + "epoch": 1.2975098296199215, + "grad_norm": 1.7308969497680664, + "learning_rate": 4.9865546504659063e-05, + "loss": 0.5517, + "num_input_tokens_seen": 1329040, + "step": 4950 + }, + { + "epoch": 1.2988204456094365, + "grad_norm": 1.425145149230957, + "learning_rate": 4.986435930953338e-05, + "loss": 0.5076, + "num_input_tokens_seen": 1330512, + "step": 4955 + }, + { + "epoch": 1.3001310615989516, + "grad_norm": 1.1248754262924194, + "learning_rate": 4.986316691034026e-05, + "loss": 0.3415, + "num_input_tokens_seen": 1331408, + "step": 4960 + }, + { + "epoch": 1.3014416775884665, + "grad_norm": 1.1896079778671265, + "learning_rate": 4.986196930732926e-05, + "loss": 0.6227, + "num_input_tokens_seen": 1332512, + "step": 4965 + }, + { + "epoch": 1.3027522935779816, + "grad_norm": 0.7298575639724731, + "learning_rate": 4.986076650075104e-05, + "loss": 0.3056, + "num_input_tokens_seen": 1334048, + "step": 4970 + }, + { + "epoch": 1.3040629095674967, + "grad_norm": 1.5282249450683594, + "learning_rate": 4.985955849085735e-05, + "loss": 0.6656, + "num_input_tokens_seen": 1335168, + "step": 4975 + }, + { + "epoch": 1.3053735255570118, + "grad_norm": 0.6302909255027771, + "learning_rate": 4.9858345277901034e-05, + "loss": 0.3446, + "num_input_tokens_seen": 1336624, + "step": 4980 + }, + { + "epoch": 1.306684141546527, + "grad_norm": 1.6390272378921509, + "learning_rate": 4.9857126862135984e-05, + "loss": 0.4584, + "num_input_tokens_seen": 1337728, + "step": 4985 + }, + { + "epoch": 1.307994757536042, + "grad_norm": 1.1867345571517944, + "learning_rate": 4.985590324381724e-05, + "loss": 0.4533, + "num_input_tokens_seen": 1339136, + "step": 4990 + }, + { + "epoch": 1.309305373525557, + "grad_norm": 1.3921595811843872, + "learning_rate": 4.98546744232009e-05, + "loss": 0.6284, + "num_input_tokens_seen": 1340784, + "step": 4995 + }, + { + "epoch": 1.310615989515072, + "grad_norm": 1.7080544233322144, + "learning_rate": 4.9853440400544136e-05, + "loss": 0.3178, + "num_input_tokens_seen": 1342384, + "step": 5000 + }, + { + "epoch": 1.311926605504587, + "grad_norm": 1.1392123699188232, + "learning_rate": 4.985220117610525e-05, + "loss": 0.4189, + "num_input_tokens_seen": 1343952, + "step": 5005 + }, + { + "epoch": 1.3132372214941022, + "grad_norm": 0.7604687213897705, + "learning_rate": 4.985095675014359e-05, + "loss": 0.3092, + "num_input_tokens_seen": 1345600, + "step": 5010 + }, + { + "epoch": 1.3145478374836173, + "grad_norm": 0.7750831842422485, + "learning_rate": 4.984970712291963e-05, + "loss": 0.5543, + "num_input_tokens_seen": 1347216, + "step": 5015 + }, + { + "epoch": 1.3158584534731324, + "grad_norm": 0.49670901894569397, + "learning_rate": 4.98484522946949e-05, + "loss": 0.4412, + "num_input_tokens_seen": 1348368, + "step": 5020 + }, + { + "epoch": 1.3171690694626474, + "grad_norm": 1.2844245433807373, + "learning_rate": 4.984719226573205e-05, + "loss": 0.4727, + "num_input_tokens_seen": 1349504, + "step": 5025 + }, + { + "epoch": 1.3184796854521625, + "grad_norm": 1.3889284133911133, + "learning_rate": 4.984592703629478e-05, + "loss": 0.534, + "num_input_tokens_seen": 1350688, + "step": 5030 + }, + { + "epoch": 1.3197903014416776, + "grad_norm": 0.8186847567558289, + "learning_rate": 4.9844656606647924e-05, + "loss": 0.5981, + "num_input_tokens_seen": 1352160, + "step": 5035 + }, + { + "epoch": 1.3211009174311927, + "grad_norm": 1.5615049600601196, + "learning_rate": 4.984338097705736e-05, + "loss": 0.3938, + "num_input_tokens_seen": 1353504, + "step": 5040 + }, + { + "epoch": 1.3224115334207078, + "grad_norm": 0.4596109986305237, + "learning_rate": 4.984210014779008e-05, + "loss": 0.4694, + "num_input_tokens_seen": 1354992, + "step": 5045 + }, + { + "epoch": 1.323722149410223, + "grad_norm": 5.876311302185059, + "learning_rate": 4.984081411911417e-05, + "loss": 0.5513, + "num_input_tokens_seen": 1355984, + "step": 5050 + }, + { + "epoch": 1.3250327653997378, + "grad_norm": 0.6398894190788269, + "learning_rate": 4.9839522891298784e-05, + "loss": 0.4361, + "num_input_tokens_seen": 1357360, + "step": 5055 + }, + { + "epoch": 1.326343381389253, + "grad_norm": 1.1669774055480957, + "learning_rate": 4.983822646461417e-05, + "loss": 0.5139, + "num_input_tokens_seen": 1358416, + "step": 5060 + }, + { + "epoch": 1.327653997378768, + "grad_norm": 0.6078435778617859, + "learning_rate": 4.9836924839331686e-05, + "loss": 0.6348, + "num_input_tokens_seen": 1359824, + "step": 5065 + }, + { + "epoch": 1.328964613368283, + "grad_norm": 1.2247501611709595, + "learning_rate": 4.983561801572374e-05, + "loss": 0.5004, + "num_input_tokens_seen": 1361056, + "step": 5070 + }, + { + "epoch": 1.3302752293577982, + "grad_norm": 1.0199707746505737, + "learning_rate": 4.983430599406386e-05, + "loss": 0.5481, + "num_input_tokens_seen": 1362048, + "step": 5075 + }, + { + "epoch": 1.3315858453473133, + "grad_norm": 0.6803602576255798, + "learning_rate": 4.983298877462664e-05, + "loss": 0.3545, + "num_input_tokens_seen": 1362960, + "step": 5080 + }, + { + "epoch": 1.3328964613368284, + "grad_norm": 0.5220121741294861, + "learning_rate": 4.983166635768778e-05, + "loss": 0.3399, + "num_input_tokens_seen": 1364192, + "step": 5085 + }, + { + "epoch": 1.3342070773263432, + "grad_norm": 0.5016363859176636, + "learning_rate": 4.9830338743524064e-05, + "loss": 0.459, + "num_input_tokens_seen": 1366240, + "step": 5090 + }, + { + "epoch": 1.3355176933158583, + "grad_norm": 1.0526654720306396, + "learning_rate": 4.982900593241334e-05, + "loss": 0.4576, + "num_input_tokens_seen": 1367584, + "step": 5095 + }, + { + "epoch": 1.3368283093053734, + "grad_norm": 1.1909514665603638, + "learning_rate": 4.982766792463458e-05, + "loss": 0.4402, + "num_input_tokens_seen": 1369152, + "step": 5100 + }, + { + "epoch": 1.3381389252948885, + "grad_norm": 1.9691070318222046, + "learning_rate": 4.9826324720467834e-05, + "loss": 0.4047, + "num_input_tokens_seen": 1370256, + "step": 5105 + }, + { + "epoch": 1.3394495412844036, + "grad_norm": 2.494715452194214, + "learning_rate": 4.982497632019421e-05, + "loss": 0.9961, + "num_input_tokens_seen": 1373184, + "step": 5110 + }, + { + "epoch": 1.3407601572739187, + "grad_norm": 1.6154533624649048, + "learning_rate": 4.9823622724095936e-05, + "loss": 0.6547, + "num_input_tokens_seen": 1374544, + "step": 5115 + }, + { + "epoch": 1.3420707732634338, + "grad_norm": 1.3831963539123535, + "learning_rate": 4.982226393245632e-05, + "loss": 0.5706, + "num_input_tokens_seen": 1375808, + "step": 5120 + }, + { + "epoch": 1.343381389252949, + "grad_norm": 1.1670019626617432, + "learning_rate": 4.982089994555975e-05, + "loss": 0.5374, + "num_input_tokens_seen": 1377216, + "step": 5125 + }, + { + "epoch": 1.344692005242464, + "grad_norm": 1.5134750604629517, + "learning_rate": 4.981953076369171e-05, + "loss": 0.537, + "num_input_tokens_seen": 1378448, + "step": 5130 + }, + { + "epoch": 1.3460026212319791, + "grad_norm": 0.7787836194038391, + "learning_rate": 4.981815638713877e-05, + "loss": 0.6506, + "num_input_tokens_seen": 1379728, + "step": 5135 + }, + { + "epoch": 1.3473132372214942, + "grad_norm": 1.5616891384124756, + "learning_rate": 4.981677681618858e-05, + "loss": 0.6218, + "num_input_tokens_seen": 1380880, + "step": 5140 + }, + { + "epoch": 1.3486238532110093, + "grad_norm": 1.6383439302444458, + "learning_rate": 4.981539205112988e-05, + "loss": 0.7655, + "num_input_tokens_seen": 1382672, + "step": 5145 + }, + { + "epoch": 1.3499344692005242, + "grad_norm": 1.7050341367721558, + "learning_rate": 4.98140020922525e-05, + "loss": 0.5043, + "num_input_tokens_seen": 1384176, + "step": 5150 + }, + { + "epoch": 1.3512450851900393, + "grad_norm": 1.2536689043045044, + "learning_rate": 4.9812606939847356e-05, + "loss": 0.5239, + "num_input_tokens_seen": 1385440, + "step": 5155 + }, + { + "epoch": 1.3525557011795544, + "grad_norm": 0.8866219520568848, + "learning_rate": 4.981120659420646e-05, + "loss": 0.5464, + "num_input_tokens_seen": 1387040, + "step": 5160 + }, + { + "epoch": 1.3538663171690695, + "grad_norm": 0.6089127659797668, + "learning_rate": 4.9809801055622887e-05, + "loss": 0.5119, + "num_input_tokens_seen": 1388048, + "step": 5165 + }, + { + "epoch": 1.3551769331585846, + "grad_norm": 0.8607680797576904, + "learning_rate": 4.980839032439082e-05, + "loss": 0.6121, + "num_input_tokens_seen": 1389072, + "step": 5170 + }, + { + "epoch": 1.3564875491480997, + "grad_norm": 1.206417202949524, + "learning_rate": 4.980697440080553e-05, + "loss": 0.4084, + "num_input_tokens_seen": 1390720, + "step": 5175 + }, + { + "epoch": 1.3577981651376148, + "grad_norm": 0.6777510643005371, + "learning_rate": 4.980555328516335e-05, + "loss": 0.2487, + "num_input_tokens_seen": 1392096, + "step": 5180 + }, + { + "epoch": 1.3591087811271296, + "grad_norm": 0.7849614024162292, + "learning_rate": 4.9804126977761747e-05, + "loss": 0.3952, + "num_input_tokens_seen": 1393376, + "step": 5185 + }, + { + "epoch": 1.3604193971166447, + "grad_norm": 0.7678700089454651, + "learning_rate": 4.980269547889921e-05, + "loss": 0.5265, + "num_input_tokens_seen": 1394704, + "step": 5190 + }, + { + "epoch": 1.3617300131061598, + "grad_norm": 1.0198034048080444, + "learning_rate": 4.980125878887537e-05, + "loss": 0.6168, + "num_input_tokens_seen": 1395888, + "step": 5195 + }, + { + "epoch": 1.363040629095675, + "grad_norm": 0.8371008634567261, + "learning_rate": 4.979981690799092e-05, + "loss": 0.6688, + "num_input_tokens_seen": 1397280, + "step": 5200 + }, + { + "epoch": 1.36435124508519, + "grad_norm": 1.8928186893463135, + "learning_rate": 4.9798369836547644e-05, + "loss": 0.4383, + "num_input_tokens_seen": 1398320, + "step": 5205 + }, + { + "epoch": 1.365661861074705, + "grad_norm": 1.0951712131500244, + "learning_rate": 4.97969175748484e-05, + "loss": 0.4469, + "num_input_tokens_seen": 1399632, + "step": 5210 + }, + { + "epoch": 1.3669724770642202, + "grad_norm": 1.298534631729126, + "learning_rate": 4.9795460123197166e-05, + "loss": 0.5019, + "num_input_tokens_seen": 1400768, + "step": 5215 + }, + { + "epoch": 1.3682830930537353, + "grad_norm": 2.1071693897247314, + "learning_rate": 4.979399748189896e-05, + "loss": 0.4145, + "num_input_tokens_seen": 1402352, + "step": 5220 + }, + { + "epoch": 1.3695937090432504, + "grad_norm": 0.46870890259742737, + "learning_rate": 4.979252965125993e-05, + "loss": 0.3216, + "num_input_tokens_seen": 1403328, + "step": 5225 + }, + { + "epoch": 1.3709043250327655, + "grad_norm": 1.0364059209823608, + "learning_rate": 4.9791056631587276e-05, + "loss": 0.5544, + "num_input_tokens_seen": 1404592, + "step": 5230 + }, + { + "epoch": 1.3722149410222806, + "grad_norm": 0.6390476226806641, + "learning_rate": 4.978957842318931e-05, + "loss": 0.474, + "num_input_tokens_seen": 1405840, + "step": 5235 + }, + { + "epoch": 1.3735255570117955, + "grad_norm": 1.2530068159103394, + "learning_rate": 4.978809502637541e-05, + "loss": 0.5589, + "num_input_tokens_seen": 1407248, + "step": 5240 + }, + { + "epoch": 1.3748361730013106, + "grad_norm": 0.9255819320678711, + "learning_rate": 4.9786606441456054e-05, + "loss": 0.4699, + "num_input_tokens_seen": 1408528, + "step": 5245 + }, + { + "epoch": 1.3761467889908257, + "grad_norm": 1.0726855993270874, + "learning_rate": 4.9785112668742796e-05, + "loss": 0.3609, + "num_input_tokens_seen": 1409792, + "step": 5250 + }, + { + "epoch": 1.3774574049803407, + "grad_norm": 2.196840524673462, + "learning_rate": 4.978361370854828e-05, + "loss": 0.4464, + "num_input_tokens_seen": 1411168, + "step": 5255 + }, + { + "epoch": 1.3787680209698558, + "grad_norm": 0.8073883652687073, + "learning_rate": 4.978210956118624e-05, + "loss": 0.4815, + "num_input_tokens_seen": 1412512, + "step": 5260 + }, + { + "epoch": 1.380078636959371, + "grad_norm": 0.6817901730537415, + "learning_rate": 4.9780600226971486e-05, + "loss": 0.5601, + "num_input_tokens_seen": 1413584, + "step": 5265 + }, + { + "epoch": 1.381389252948886, + "grad_norm": 0.5656224489212036, + "learning_rate": 4.977908570621993e-05, + "loss": 0.6072, + "num_input_tokens_seen": 1415520, + "step": 5270 + }, + { + "epoch": 1.382699868938401, + "grad_norm": 0.8602381348609924, + "learning_rate": 4.977756599924854e-05, + "loss": 0.5795, + "num_input_tokens_seen": 1416576, + "step": 5275 + }, + { + "epoch": 1.384010484927916, + "grad_norm": 0.9414363503456116, + "learning_rate": 4.97760411063754e-05, + "loss": 0.3491, + "num_input_tokens_seen": 1418032, + "step": 5280 + }, + { + "epoch": 1.385321100917431, + "grad_norm": 0.6881532073020935, + "learning_rate": 4.977451102791968e-05, + "loss": 0.5349, + "num_input_tokens_seen": 1419264, + "step": 5285 + }, + { + "epoch": 1.3866317169069462, + "grad_norm": 2.290332317352295, + "learning_rate": 4.977297576420159e-05, + "loss": 0.5829, + "num_input_tokens_seen": 1420224, + "step": 5290 + }, + { + "epoch": 1.3879423328964613, + "grad_norm": 2.5132477283477783, + "learning_rate": 4.977143531554249e-05, + "loss": 0.5434, + "num_input_tokens_seen": 1421552, + "step": 5295 + }, + { + "epoch": 1.3892529488859764, + "grad_norm": 0.5175978541374207, + "learning_rate": 4.9769889682264774e-05, + "loss": 0.5536, + "num_input_tokens_seen": 1422832, + "step": 5300 + }, + { + "epoch": 1.3905635648754915, + "grad_norm": 0.8095912337303162, + "learning_rate": 4.976833886469196e-05, + "loss": 0.4924, + "num_input_tokens_seen": 1423952, + "step": 5305 + }, + { + "epoch": 1.3918741808650066, + "grad_norm": 0.8183461427688599, + "learning_rate": 4.97667828631486e-05, + "loss": 0.3191, + "num_input_tokens_seen": 1425168, + "step": 5310 + }, + { + "epoch": 1.3931847968545217, + "grad_norm": 1.145223617553711, + "learning_rate": 4.976522167796038e-05, + "loss": 0.5516, + "num_input_tokens_seen": 1426160, + "step": 5315 + }, + { + "epoch": 1.3944954128440368, + "grad_norm": 0.8545402884483337, + "learning_rate": 4.9763655309454064e-05, + "loss": 0.5665, + "num_input_tokens_seen": 1427440, + "step": 5320 + }, + { + "epoch": 1.3958060288335519, + "grad_norm": 0.46477702260017395, + "learning_rate": 4.976208375795748e-05, + "loss": 0.4709, + "num_input_tokens_seen": 1428496, + "step": 5325 + }, + { + "epoch": 1.397116644823067, + "grad_norm": 1.5768555402755737, + "learning_rate": 4.9760507023799544e-05, + "loss": 0.4407, + "num_input_tokens_seen": 1429568, + "step": 5330 + }, + { + "epoch": 1.3984272608125818, + "grad_norm": 0.9834166765213013, + "learning_rate": 4.9758925107310276e-05, + "loss": 0.5139, + "num_input_tokens_seen": 1430800, + "step": 5335 + }, + { + "epoch": 1.399737876802097, + "grad_norm": 1.7082493305206299, + "learning_rate": 4.975733800882077e-05, + "loss": 0.5121, + "num_input_tokens_seen": 1431968, + "step": 5340 + }, + { + "epoch": 1.401048492791612, + "grad_norm": 3.1272220611572266, + "learning_rate": 4.975574572866318e-05, + "loss": 0.6382, + "num_input_tokens_seen": 1433440, + "step": 5345 + }, + { + "epoch": 1.4023591087811271, + "grad_norm": 0.8326922655105591, + "learning_rate": 4.97541482671708e-05, + "loss": 0.3873, + "num_input_tokens_seen": 1434640, + "step": 5350 + }, + { + "epoch": 1.4036697247706422, + "grad_norm": 0.6369110345840454, + "learning_rate": 4.975254562467794e-05, + "loss": 0.4993, + "num_input_tokens_seen": 1436224, + "step": 5355 + }, + { + "epoch": 1.4049803407601573, + "grad_norm": 0.3529784679412842, + "learning_rate": 4.9750937801520064e-05, + "loss": 0.4075, + "num_input_tokens_seen": 1437136, + "step": 5360 + }, + { + "epoch": 1.4062909567496724, + "grad_norm": 1.2307825088500977, + "learning_rate": 4.974932479803367e-05, + "loss": 0.5732, + "num_input_tokens_seen": 1438384, + "step": 5365 + }, + { + "epoch": 1.4076015727391873, + "grad_norm": 0.9378088116645813, + "learning_rate": 4.974770661455636e-05, + "loss": 0.6208, + "num_input_tokens_seen": 1439600, + "step": 5370 + }, + { + "epoch": 1.4089121887287024, + "grad_norm": 0.670146644115448, + "learning_rate": 4.974608325142681e-05, + "loss": 0.5294, + "num_input_tokens_seen": 1440768, + "step": 5375 + }, + { + "epoch": 1.4102228047182175, + "grad_norm": 0.700986921787262, + "learning_rate": 4.97444547089848e-05, + "loss": 0.4458, + "num_input_tokens_seen": 1442320, + "step": 5380 + }, + { + "epoch": 1.4115334207077326, + "grad_norm": 2.0755813121795654, + "learning_rate": 4.974282098757118e-05, + "loss": 0.4834, + "num_input_tokens_seen": 1443456, + "step": 5385 + }, + { + "epoch": 1.4128440366972477, + "grad_norm": 0.7429225444793701, + "learning_rate": 4.974118208752787e-05, + "loss": 0.4132, + "num_input_tokens_seen": 1444864, + "step": 5390 + }, + { + "epoch": 1.4141546526867628, + "grad_norm": 0.7103058099746704, + "learning_rate": 4.97395380091979e-05, + "loss": 0.4338, + "num_input_tokens_seen": 1446176, + "step": 5395 + }, + { + "epoch": 1.4154652686762779, + "grad_norm": 0.8550195693969727, + "learning_rate": 4.9737888752925375e-05, + "loss": 0.4606, + "num_input_tokens_seen": 1447376, + "step": 5400 + }, + { + "epoch": 1.416775884665793, + "grad_norm": 1.508304476737976, + "learning_rate": 4.973623431905548e-05, + "loss": 0.5389, + "num_input_tokens_seen": 1448368, + "step": 5405 + }, + { + "epoch": 1.418086500655308, + "grad_norm": 1.0948044061660767, + "learning_rate": 4.973457470793447e-05, + "loss": 0.5465, + "num_input_tokens_seen": 1449456, + "step": 5410 + }, + { + "epoch": 1.4193971166448232, + "grad_norm": 0.6115325689315796, + "learning_rate": 4.973290991990972e-05, + "loss": 0.4601, + "num_input_tokens_seen": 1451088, + "step": 5415 + }, + { + "epoch": 1.4207077326343382, + "grad_norm": 1.1457611322402954, + "learning_rate": 4.973123995532966e-05, + "loss": 0.5225, + "num_input_tokens_seen": 1452240, + "step": 5420 + }, + { + "epoch": 1.4220183486238533, + "grad_norm": 0.8227062225341797, + "learning_rate": 4.972956481454381e-05, + "loss": 0.4726, + "num_input_tokens_seen": 1453648, + "step": 5425 + }, + { + "epoch": 1.4233289646133682, + "grad_norm": 0.9384662508964539, + "learning_rate": 4.972788449790277e-05, + "loss": 0.5678, + "num_input_tokens_seen": 1455584, + "step": 5430 + }, + { + "epoch": 1.4246395806028833, + "grad_norm": 0.7072715759277344, + "learning_rate": 4.9726199005758234e-05, + "loss": 0.3849, + "num_input_tokens_seen": 1457408, + "step": 5435 + }, + { + "epoch": 1.4259501965923984, + "grad_norm": 0.8595460057258606, + "learning_rate": 4.972450833846297e-05, + "loss": 0.4865, + "num_input_tokens_seen": 1458704, + "step": 5440 + }, + { + "epoch": 1.4272608125819135, + "grad_norm": 1.0292150974273682, + "learning_rate": 4.972281249637083e-05, + "loss": 0.3492, + "num_input_tokens_seen": 1461232, + "step": 5445 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.7816736102104187, + "learning_rate": 4.9721111479836756e-05, + "loss": 0.5126, + "num_input_tokens_seen": 1462768, + "step": 5450 + }, + { + "epoch": 1.4298820445609437, + "grad_norm": 1.1008294820785522, + "learning_rate": 4.971940528921676e-05, + "loss": 0.4543, + "num_input_tokens_seen": 1464160, + "step": 5455 + }, + { + "epoch": 1.4311926605504588, + "grad_norm": 1.2781497240066528, + "learning_rate": 4.9717693924867944e-05, + "loss": 0.4163, + "num_input_tokens_seen": 1465264, + "step": 5460 + }, + { + "epoch": 1.4325032765399737, + "grad_norm": 0.9671511650085449, + "learning_rate": 4.9715977387148504e-05, + "loss": 0.4108, + "num_input_tokens_seen": 1466592, + "step": 5465 + }, + { + "epoch": 1.4338138925294888, + "grad_norm": 1.0626291036605835, + "learning_rate": 4.971425567641771e-05, + "loss": 0.5781, + "num_input_tokens_seen": 1467744, + "step": 5470 + }, + { + "epoch": 1.4351245085190039, + "grad_norm": 0.8136863708496094, + "learning_rate": 4.97125287930359e-05, + "loss": 0.4171, + "num_input_tokens_seen": 1468768, + "step": 5475 + }, + { + "epoch": 1.436435124508519, + "grad_norm": 2.2941901683807373, + "learning_rate": 4.971079673736451e-05, + "loss": 0.4955, + "num_input_tokens_seen": 1470144, + "step": 5480 + }, + { + "epoch": 1.437745740498034, + "grad_norm": 0.49133917689323425, + "learning_rate": 4.970905950976606e-05, + "loss": 0.6547, + "num_input_tokens_seen": 1471488, + "step": 5485 + }, + { + "epoch": 1.4390563564875491, + "grad_norm": 2.4882912635803223, + "learning_rate": 4.970731711060415e-05, + "loss": 0.7066, + "num_input_tokens_seen": 1472784, + "step": 5490 + }, + { + "epoch": 1.4403669724770642, + "grad_norm": 0.7773187756538391, + "learning_rate": 4.970556954024346e-05, + "loss": 0.4282, + "num_input_tokens_seen": 1474112, + "step": 5495 + }, + { + "epoch": 1.4416775884665793, + "grad_norm": 2.077371597290039, + "learning_rate": 4.970381679904975e-05, + "loss": 0.3785, + "num_input_tokens_seen": 1475120, + "step": 5500 + }, + { + "epoch": 1.4429882044560944, + "grad_norm": 0.7187175154685974, + "learning_rate": 4.970205888738988e-05, + "loss": 0.6274, + "num_input_tokens_seen": 1476368, + "step": 5505 + }, + { + "epoch": 1.4442988204456095, + "grad_norm": 2.37835693359375, + "learning_rate": 4.970029580563176e-05, + "loss": 0.4744, + "num_input_tokens_seen": 1477760, + "step": 5510 + }, + { + "epoch": 1.4456094364351246, + "grad_norm": 0.9618066549301147, + "learning_rate": 4.96985275541444e-05, + "loss": 0.382, + "num_input_tokens_seen": 1479136, + "step": 5515 + }, + { + "epoch": 1.4469200524246395, + "grad_norm": 1.155692219734192, + "learning_rate": 4.969675413329791e-05, + "loss": 0.3317, + "num_input_tokens_seen": 1480048, + "step": 5520 + }, + { + "epoch": 1.4482306684141546, + "grad_norm": 0.4981410801410675, + "learning_rate": 4.969497554346344e-05, + "loss": 0.4033, + "num_input_tokens_seen": 1481360, + "step": 5525 + }, + { + "epoch": 1.4495412844036697, + "grad_norm": 0.9506641030311584, + "learning_rate": 4.969319178501327e-05, + "loss": 0.5941, + "num_input_tokens_seen": 1482864, + "step": 5530 + }, + { + "epoch": 1.4508519003931848, + "grad_norm": 1.0454243421554565, + "learning_rate": 4.969140285832072e-05, + "loss": 0.3464, + "num_input_tokens_seen": 1484336, + "step": 5535 + }, + { + "epoch": 1.4521625163826999, + "grad_norm": 0.8238692879676819, + "learning_rate": 4.968960876376022e-05, + "loss": 0.4356, + "num_input_tokens_seen": 1485472, + "step": 5540 + }, + { + "epoch": 1.453473132372215, + "grad_norm": 0.5614863634109497, + "learning_rate": 4.9687809501707264e-05, + "loss": 0.4283, + "num_input_tokens_seen": 1487264, + "step": 5545 + }, + { + "epoch": 1.45478374836173, + "grad_norm": 0.776120126247406, + "learning_rate": 4.968600507253843e-05, + "loss": 0.4804, + "num_input_tokens_seen": 1488400, + "step": 5550 + }, + { + "epoch": 1.456094364351245, + "grad_norm": 0.9073880314826965, + "learning_rate": 4.968419547663139e-05, + "loss": 0.3878, + "num_input_tokens_seen": 1489424, + "step": 5555 + }, + { + "epoch": 1.45740498034076, + "grad_norm": 1.4581652879714966, + "learning_rate": 4.9682380714364897e-05, + "loss": 0.5697, + "num_input_tokens_seen": 1490320, + "step": 5560 + }, + { + "epoch": 1.4587155963302751, + "grad_norm": 0.6152535676956177, + "learning_rate": 4.968056078611876e-05, + "loss": 0.5917, + "num_input_tokens_seen": 1492640, + "step": 5565 + }, + { + "epoch": 1.4600262123197902, + "grad_norm": 1.158154010772705, + "learning_rate": 4.96787356922739e-05, + "loss": 0.5868, + "num_input_tokens_seen": 1493872, + "step": 5570 + }, + { + "epoch": 1.4613368283093053, + "grad_norm": 0.9468458294868469, + "learning_rate": 4.9676905433212296e-05, + "loss": 0.5079, + "num_input_tokens_seen": 1495216, + "step": 5575 + }, + { + "epoch": 1.4626474442988204, + "grad_norm": 1.3104547262191772, + "learning_rate": 4.967507000931702e-05, + "loss": 0.4804, + "num_input_tokens_seen": 1496480, + "step": 5580 + }, + { + "epoch": 1.4639580602883355, + "grad_norm": 0.6972042918205261, + "learning_rate": 4.967322942097225e-05, + "loss": 0.4418, + "num_input_tokens_seen": 1497936, + "step": 5585 + }, + { + "epoch": 1.4652686762778506, + "grad_norm": 1.3180348873138428, + "learning_rate": 4.9671383668563167e-05, + "loss": 0.454, + "num_input_tokens_seen": 1499280, + "step": 5590 + }, + { + "epoch": 1.4665792922673657, + "grad_norm": 0.6575278639793396, + "learning_rate": 4.966953275247612e-05, + "loss": 0.5505, + "num_input_tokens_seen": 1500464, + "step": 5595 + }, + { + "epoch": 1.4678899082568808, + "grad_norm": 0.522504985332489, + "learning_rate": 4.966767667309849e-05, + "loss": 0.3442, + "num_input_tokens_seen": 1501824, + "step": 5600 + }, + { + "epoch": 1.469200524246396, + "grad_norm": 0.6469646096229553, + "learning_rate": 4.966581543081876e-05, + "loss": 0.4907, + "num_input_tokens_seen": 1503024, + "step": 5605 + }, + { + "epoch": 1.470511140235911, + "grad_norm": 3.4661996364593506, + "learning_rate": 4.966394902602649e-05, + "loss": 0.631, + "num_input_tokens_seen": 1504800, + "step": 5610 + }, + { + "epoch": 1.4718217562254259, + "grad_norm": 2.267263889312744, + "learning_rate": 4.966207745911229e-05, + "loss": 0.4033, + "num_input_tokens_seen": 1506016, + "step": 5615 + }, + { + "epoch": 1.473132372214941, + "grad_norm": 1.1045714616775513, + "learning_rate": 4.9660200730467894e-05, + "loss": 0.6106, + "num_input_tokens_seen": 1507360, + "step": 5620 + }, + { + "epoch": 1.474442988204456, + "grad_norm": 0.9895575046539307, + "learning_rate": 4.9658318840486095e-05, + "loss": 0.5802, + "num_input_tokens_seen": 1508528, + "step": 5625 + }, + { + "epoch": 1.4757536041939712, + "grad_norm": 2.147512674331665, + "learning_rate": 4.9656431789560765e-05, + "loss": 0.5152, + "num_input_tokens_seen": 1509584, + "step": 5630 + }, + { + "epoch": 1.4770642201834863, + "grad_norm": 1.3974361419677734, + "learning_rate": 4.9654539578086865e-05, + "loss": 0.8406, + "num_input_tokens_seen": 1511008, + "step": 5635 + }, + { + "epoch": 1.4783748361730014, + "grad_norm": 0.8087236881256104, + "learning_rate": 4.965264220646044e-05, + "loss": 0.532, + "num_input_tokens_seen": 1512448, + "step": 5640 + }, + { + "epoch": 1.4796854521625165, + "grad_norm": 1.6618154048919678, + "learning_rate": 4.965073967507859e-05, + "loss": 0.526, + "num_input_tokens_seen": 1513712, + "step": 5645 + }, + { + "epoch": 1.4809960681520313, + "grad_norm": 1.1800471544265747, + "learning_rate": 4.9648831984339525e-05, + "loss": 0.6023, + "num_input_tokens_seen": 1515632, + "step": 5650 + }, + { + "epoch": 1.4823066841415464, + "grad_norm": 1.1400423049926758, + "learning_rate": 4.9646919134642513e-05, + "loss": 0.3562, + "num_input_tokens_seen": 1516592, + "step": 5655 + }, + { + "epoch": 1.4836173001310615, + "grad_norm": 1.086679220199585, + "learning_rate": 4.964500112638791e-05, + "loss": 0.4619, + "num_input_tokens_seen": 1518080, + "step": 5660 + }, + { + "epoch": 1.4849279161205766, + "grad_norm": 1.156542420387268, + "learning_rate": 4.964307795997716e-05, + "loss": 0.3543, + "num_input_tokens_seen": 1519344, + "step": 5665 + }, + { + "epoch": 1.4862385321100917, + "grad_norm": 2.199963092803955, + "learning_rate": 4.964114963581277e-05, + "loss": 0.4681, + "num_input_tokens_seen": 1520384, + "step": 5670 + }, + { + "epoch": 1.4875491480996068, + "grad_norm": 1.7003588676452637, + "learning_rate": 4.963921615429833e-05, + "loss": 0.5288, + "num_input_tokens_seen": 1521840, + "step": 5675 + }, + { + "epoch": 1.488859764089122, + "grad_norm": 0.8601107001304626, + "learning_rate": 4.963727751583853e-05, + "loss": 0.5571, + "num_input_tokens_seen": 1522960, + "step": 5680 + }, + { + "epoch": 1.490170380078637, + "grad_norm": 0.5830681324005127, + "learning_rate": 4.963533372083912e-05, + "loss": 0.4615, + "num_input_tokens_seen": 1524416, + "step": 5685 + }, + { + "epoch": 1.491480996068152, + "grad_norm": 1.7958369255065918, + "learning_rate": 4.963338476970692e-05, + "loss": 0.5468, + "num_input_tokens_seen": 1525808, + "step": 5690 + }, + { + "epoch": 1.4927916120576672, + "grad_norm": 1.8775278329849243, + "learning_rate": 4.963143066284986e-05, + "loss": 0.3973, + "num_input_tokens_seen": 1527088, + "step": 5695 + }, + { + "epoch": 1.4941022280471823, + "grad_norm": 2.973194122314453, + "learning_rate": 4.962947140067691e-05, + "loss": 0.5338, + "num_input_tokens_seen": 1528432, + "step": 5700 + }, + { + "epoch": 1.4954128440366974, + "grad_norm": 0.7447792291641235, + "learning_rate": 4.962750698359816e-05, + "loss": 0.6326, + "num_input_tokens_seen": 1529520, + "step": 5705 + }, + { + "epoch": 1.4967234600262123, + "grad_norm": 0.7448498606681824, + "learning_rate": 4.962553741202475e-05, + "loss": 0.4788, + "num_input_tokens_seen": 1530720, + "step": 5710 + }, + { + "epoch": 1.4980340760157274, + "grad_norm": 1.3587487936019897, + "learning_rate": 4.962356268636891e-05, + "loss": 0.5605, + "num_input_tokens_seen": 1531952, + "step": 5715 + }, + { + "epoch": 1.4993446920052425, + "grad_norm": 0.7149798274040222, + "learning_rate": 4.962158280704395e-05, + "loss": 0.4662, + "num_input_tokens_seen": 1533456, + "step": 5720 + }, + { + "epoch": 1.5003931847968546, + "eval_loss": 0.5115170478820801, + "eval_runtime": 16.869, + "eval_samples_per_second": 50.27, + "eval_steps_per_second": 25.135, + "num_input_tokens_seen": 1534400, + "step": 5724 + }, + { + "epoch": 1.5006553079947575, + "grad_norm": 0.731968343257904, + "learning_rate": 4.961959777446425e-05, + "loss": 0.5307, + "num_input_tokens_seen": 1534720, + "step": 5725 + }, + { + "epoch": 1.5019659239842726, + "grad_norm": 0.7347331047058105, + "learning_rate": 4.961760758904527e-05, + "loss": 0.3311, + "num_input_tokens_seen": 1536992, + "step": 5730 + }, + { + "epoch": 1.5032765399737875, + "grad_norm": 1.2266509532928467, + "learning_rate": 4.961561225120356e-05, + "loss": 0.6005, + "num_input_tokens_seen": 1538288, + "step": 5735 + }, + { + "epoch": 1.5045871559633026, + "grad_norm": 0.7551884055137634, + "learning_rate": 4.9613611761356736e-05, + "loss": 0.4807, + "num_input_tokens_seen": 1539296, + "step": 5740 + }, + { + "epoch": 1.5058977719528177, + "grad_norm": 0.5654918551445007, + "learning_rate": 4.96116061199235e-05, + "loss": 0.4038, + "num_input_tokens_seen": 1540896, + "step": 5745 + }, + { + "epoch": 1.5072083879423328, + "grad_norm": 1.062219262123108, + "learning_rate": 4.960959532732363e-05, + "loss": 0.433, + "num_input_tokens_seen": 1542000, + "step": 5750 + }, + { + "epoch": 1.508519003931848, + "grad_norm": 0.7273585200309753, + "learning_rate": 4.9607579383977984e-05, + "loss": 0.4251, + "num_input_tokens_seen": 1543536, + "step": 5755 + }, + { + "epoch": 1.509829619921363, + "grad_norm": 1.5300326347351074, + "learning_rate": 4.9605558290308484e-05, + "loss": 0.5783, + "num_input_tokens_seen": 1545056, + "step": 5760 + }, + { + "epoch": 1.511140235910878, + "grad_norm": 2.6009697914123535, + "learning_rate": 4.9603532046738154e-05, + "loss": 0.4075, + "num_input_tokens_seen": 1546128, + "step": 5765 + }, + { + "epoch": 1.5124508519003932, + "grad_norm": 1.0283674001693726, + "learning_rate": 4.960150065369108e-05, + "loss": 0.3033, + "num_input_tokens_seen": 1547536, + "step": 5770 + }, + { + "epoch": 1.5137614678899083, + "grad_norm": 0.47772708535194397, + "learning_rate": 4.9599464111592417e-05, + "loss": 0.4794, + "num_input_tokens_seen": 1549152, + "step": 5775 + }, + { + "epoch": 1.5150720838794234, + "grad_norm": 0.5092713832855225, + "learning_rate": 4.959742242086843e-05, + "loss": 0.4688, + "num_input_tokens_seen": 1550464, + "step": 5780 + }, + { + "epoch": 1.5163826998689385, + "grad_norm": 0.6350592970848083, + "learning_rate": 4.959537558194643e-05, + "loss": 0.4473, + "num_input_tokens_seen": 1551808, + "step": 5785 + }, + { + "epoch": 1.5176933158584536, + "grad_norm": 0.9914599657058716, + "learning_rate": 4.959332359525482e-05, + "loss": 0.4709, + "num_input_tokens_seen": 1553056, + "step": 5790 + }, + { + "epoch": 1.5190039318479687, + "grad_norm": 0.9375805854797363, + "learning_rate": 4.9591266461223075e-05, + "loss": 0.4694, + "num_input_tokens_seen": 1554032, + "step": 5795 + }, + { + "epoch": 1.5203145478374838, + "grad_norm": 0.8262470960617065, + "learning_rate": 4.958920418028175e-05, + "loss": 0.3603, + "num_input_tokens_seen": 1555360, + "step": 5800 + }, + { + "epoch": 1.5216251638269986, + "grad_norm": 1.0025360584259033, + "learning_rate": 4.958713675286247e-05, + "loss": 0.513, + "num_input_tokens_seen": 1556384, + "step": 5805 + }, + { + "epoch": 1.5229357798165137, + "grad_norm": 0.8986231684684753, + "learning_rate": 4.9585064179397955e-05, + "loss": 0.4739, + "num_input_tokens_seen": 1557968, + "step": 5810 + }, + { + "epoch": 1.5242463958060288, + "grad_norm": 1.3707995414733887, + "learning_rate": 4.958298646032199e-05, + "loss": 0.571, + "num_input_tokens_seen": 1559056, + "step": 5815 + }, + { + "epoch": 1.525557011795544, + "grad_norm": 1.5646862983703613, + "learning_rate": 4.958090359606943e-05, + "loss": 0.4626, + "num_input_tokens_seen": 1560336, + "step": 5820 + }, + { + "epoch": 1.526867627785059, + "grad_norm": 1.3486241102218628, + "learning_rate": 4.957881558707622e-05, + "loss": 0.3891, + "num_input_tokens_seen": 1561344, + "step": 5825 + }, + { + "epoch": 1.528178243774574, + "grad_norm": 0.8138800263404846, + "learning_rate": 4.957672243377938e-05, + "loss": 0.542, + "num_input_tokens_seen": 1562960, + "step": 5830 + }, + { + "epoch": 1.529488859764089, + "grad_norm": 1.1160448789596558, + "learning_rate": 4.9574624136617e-05, + "loss": 0.4602, + "num_input_tokens_seen": 1564736, + "step": 5835 + }, + { + "epoch": 1.530799475753604, + "grad_norm": 1.7018227577209473, + "learning_rate": 4.9572520696028245e-05, + "loss": 0.4171, + "num_input_tokens_seen": 1566624, + "step": 5840 + }, + { + "epoch": 1.5321100917431192, + "grad_norm": 0.6432651281356812, + "learning_rate": 4.9570412112453377e-05, + "loss": 0.4428, + "num_input_tokens_seen": 1568336, + "step": 5845 + }, + { + "epoch": 1.5334207077326343, + "grad_norm": 0.505611777305603, + "learning_rate": 4.95682983863337e-05, + "loss": 0.5026, + "num_input_tokens_seen": 1570208, + "step": 5850 + }, + { + "epoch": 1.5347313237221494, + "grad_norm": 1.4866430759429932, + "learning_rate": 4.9566179518111625e-05, + "loss": 0.4748, + "num_input_tokens_seen": 1571504, + "step": 5855 + }, + { + "epoch": 1.5360419397116645, + "grad_norm": 1.7853851318359375, + "learning_rate": 4.9564055508230616e-05, + "loss": 0.4932, + "num_input_tokens_seen": 1572400, + "step": 5860 + }, + { + "epoch": 1.5373525557011796, + "grad_norm": 0.6435667276382446, + "learning_rate": 4.9561926357135235e-05, + "loss": 0.5005, + "num_input_tokens_seen": 1573856, + "step": 5865 + }, + { + "epoch": 1.5386631716906947, + "grad_norm": 0.7923941016197205, + "learning_rate": 4.9559792065271116e-05, + "loss": 0.3814, + "num_input_tokens_seen": 1574976, + "step": 5870 + }, + { + "epoch": 1.5399737876802098, + "grad_norm": 0.9514920711517334, + "learning_rate": 4.955765263308495e-05, + "loss": 0.6464, + "num_input_tokens_seen": 1576016, + "step": 5875 + }, + { + "epoch": 1.5412844036697249, + "grad_norm": 1.2906173467636108, + "learning_rate": 4.9555508061024514e-05, + "loss": 0.4023, + "num_input_tokens_seen": 1577552, + "step": 5880 + }, + { + "epoch": 1.54259501965924, + "grad_norm": 1.1289206743240356, + "learning_rate": 4.955335834953867e-05, + "loss": 0.3169, + "num_input_tokens_seen": 1578944, + "step": 5885 + }, + { + "epoch": 1.543905635648755, + "grad_norm": 1.0248501300811768, + "learning_rate": 4.955120349907736e-05, + "loss": 0.3249, + "num_input_tokens_seen": 1580336, + "step": 5890 + }, + { + "epoch": 1.5452162516382701, + "grad_norm": 0.8925690054893494, + "learning_rate": 4.954904351009157e-05, + "loss": 0.8458, + "num_input_tokens_seen": 1581456, + "step": 5895 + }, + { + "epoch": 1.546526867627785, + "grad_norm": 0.8976151347160339, + "learning_rate": 4.954687838303339e-05, + "loss": 0.4874, + "num_input_tokens_seen": 1582976, + "step": 5900 + }, + { + "epoch": 1.5478374836173001, + "grad_norm": 0.5918650031089783, + "learning_rate": 4.9544708118355984e-05, + "loss": 0.3836, + "num_input_tokens_seen": 1584544, + "step": 5905 + }, + { + "epoch": 1.5491480996068152, + "grad_norm": 0.86038738489151, + "learning_rate": 4.9542532716513576e-05, + "loss": 0.6509, + "num_input_tokens_seen": 1585936, + "step": 5910 + }, + { + "epoch": 1.5504587155963303, + "grad_norm": 0.8281438946723938, + "learning_rate": 4.9540352177961466e-05, + "loss": 0.4894, + "num_input_tokens_seen": 1587056, + "step": 5915 + }, + { + "epoch": 1.5517693315858454, + "grad_norm": 1.009172797203064, + "learning_rate": 4.953816650315606e-05, + "loss": 0.448, + "num_input_tokens_seen": 1588704, + "step": 5920 + }, + { + "epoch": 1.5530799475753603, + "grad_norm": 1.057616114616394, + "learning_rate": 4.9535975692554805e-05, + "loss": 0.3448, + "num_input_tokens_seen": 1590016, + "step": 5925 + }, + { + "epoch": 1.5543905635648754, + "grad_norm": 7.002635955810547, + "learning_rate": 4.953377974661623e-05, + "loss": 0.512, + "num_input_tokens_seen": 1591200, + "step": 5930 + }, + { + "epoch": 1.5557011795543905, + "grad_norm": 1.3114513158798218, + "learning_rate": 4.953157866579994e-05, + "loss": 0.5671, + "num_input_tokens_seen": 1592576, + "step": 5935 + }, + { + "epoch": 1.5570117955439056, + "grad_norm": 1.601340651512146, + "learning_rate": 4.9529372450566624e-05, + "loss": 0.2956, + "num_input_tokens_seen": 1593680, + "step": 5940 + }, + { + "epoch": 1.5583224115334207, + "grad_norm": 0.6375578045845032, + "learning_rate": 4.952716110137804e-05, + "loss": 0.4718, + "num_input_tokens_seen": 1594800, + "step": 5945 + }, + { + "epoch": 1.5596330275229358, + "grad_norm": 0.7867140769958496, + "learning_rate": 4.952494461869702e-05, + "loss": 0.4343, + "num_input_tokens_seen": 1596096, + "step": 5950 + }, + { + "epoch": 1.5609436435124509, + "grad_norm": 0.8495163917541504, + "learning_rate": 4.952272300298747e-05, + "loss": 0.5395, + "num_input_tokens_seen": 1597344, + "step": 5955 + }, + { + "epoch": 1.562254259501966, + "grad_norm": 1.1946884393692017, + "learning_rate": 4.952049625471436e-05, + "loss": 0.4992, + "num_input_tokens_seen": 1598864, + "step": 5960 + }, + { + "epoch": 1.563564875491481, + "grad_norm": 0.5561442971229553, + "learning_rate": 4.951826437434375e-05, + "loss": 0.432, + "num_input_tokens_seen": 1600144, + "step": 5965 + }, + { + "epoch": 1.5648754914809961, + "grad_norm": 0.31599515676498413, + "learning_rate": 4.951602736234278e-05, + "loss": 0.4207, + "num_input_tokens_seen": 1602080, + "step": 5970 + }, + { + "epoch": 1.5661861074705112, + "grad_norm": 1.9786511659622192, + "learning_rate": 4.9513785219179634e-05, + "loss": 0.4634, + "num_input_tokens_seen": 1603664, + "step": 5975 + }, + { + "epoch": 1.5674967234600263, + "grad_norm": 1.4597066640853882, + "learning_rate": 4.95115379453236e-05, + "loss": 0.5595, + "num_input_tokens_seen": 1605616, + "step": 5980 + }, + { + "epoch": 1.5688073394495414, + "grad_norm": 0.4067003130912781, + "learning_rate": 4.950928554124503e-05, + "loss": 0.4519, + "num_input_tokens_seen": 1606720, + "step": 5985 + }, + { + "epoch": 1.5701179554390565, + "grad_norm": 0.7267105579376221, + "learning_rate": 4.950702800741534e-05, + "loss": 0.834, + "num_input_tokens_seen": 1608240, + "step": 5990 + }, + { + "epoch": 1.5714285714285714, + "grad_norm": 1.4552592039108276, + "learning_rate": 4.950476534430703e-05, + "loss": 0.6621, + "num_input_tokens_seen": 1609504, + "step": 5995 + }, + { + "epoch": 1.5727391874180865, + "grad_norm": 0.8202419281005859, + "learning_rate": 4.950249755239369e-05, + "loss": 0.4699, + "num_input_tokens_seen": 1610608, + "step": 6000 + }, + { + "epoch": 1.5740498034076016, + "grad_norm": 2.787201166152954, + "learning_rate": 4.950022463214994e-05, + "loss": 0.5832, + "num_input_tokens_seen": 1611968, + "step": 6005 + }, + { + "epoch": 1.5753604193971167, + "grad_norm": 0.45927906036376953, + "learning_rate": 4.949794658405151e-05, + "loss": 0.3024, + "num_input_tokens_seen": 1613712, + "step": 6010 + }, + { + "epoch": 1.5766710353866316, + "grad_norm": 0.5736846923828125, + "learning_rate": 4.9495663408575185e-05, + "loss": 0.4357, + "num_input_tokens_seen": 1615280, + "step": 6015 + }, + { + "epoch": 1.5779816513761467, + "grad_norm": 0.4431515634059906, + "learning_rate": 4.949337510619884e-05, + "loss": 0.4228, + "num_input_tokens_seen": 1616992, + "step": 6020 + }, + { + "epoch": 1.5792922673656618, + "grad_norm": 0.6287021636962891, + "learning_rate": 4.94910816774014e-05, + "loss": 0.491, + "num_input_tokens_seen": 1618592, + "step": 6025 + }, + { + "epoch": 1.5806028833551768, + "grad_norm": 1.1063899993896484, + "learning_rate": 4.9488783122662886e-05, + "loss": 0.4701, + "num_input_tokens_seen": 1619824, + "step": 6030 + }, + { + "epoch": 1.581913499344692, + "grad_norm": 0.7941713929176331, + "learning_rate": 4.9486479442464374e-05, + "loss": 0.4997, + "num_input_tokens_seen": 1620800, + "step": 6035 + }, + { + "epoch": 1.583224115334207, + "grad_norm": 1.2041486501693726, + "learning_rate": 4.948417063728803e-05, + "loss": 0.504, + "num_input_tokens_seen": 1622000, + "step": 6040 + }, + { + "epoch": 1.5845347313237221, + "grad_norm": 0.8138375878334045, + "learning_rate": 4.948185670761707e-05, + "loss": 0.3577, + "num_input_tokens_seen": 1623328, + "step": 6045 + }, + { + "epoch": 1.5858453473132372, + "grad_norm": 1.7278834581375122, + "learning_rate": 4.94795376539358e-05, + "loss": 0.4029, + "num_input_tokens_seen": 1624432, + "step": 6050 + }, + { + "epoch": 1.5871559633027523, + "grad_norm": 2.501671314239502, + "learning_rate": 4.947721347672961e-05, + "loss": 0.4645, + "num_input_tokens_seen": 1626304, + "step": 6055 + }, + { + "epoch": 1.5884665792922674, + "grad_norm": 0.8351472020149231, + "learning_rate": 4.9474884176484924e-05, + "loss": 0.6423, + "num_input_tokens_seen": 1628480, + "step": 6060 + }, + { + "epoch": 1.5897771952817825, + "grad_norm": 1.4626615047454834, + "learning_rate": 4.947254975368927e-05, + "loss": 0.6244, + "num_input_tokens_seen": 1630048, + "step": 6065 + }, + { + "epoch": 1.5910878112712976, + "grad_norm": 0.359089195728302, + "learning_rate": 4.947021020883124e-05, + "loss": 0.4844, + "num_input_tokens_seen": 1631600, + "step": 6070 + }, + { + "epoch": 1.5923984272608127, + "grad_norm": 0.6886135935783386, + "learning_rate": 4.9467865542400494e-05, + "loss": 0.4394, + "num_input_tokens_seen": 1633232, + "step": 6075 + }, + { + "epoch": 1.5937090432503278, + "grad_norm": 0.6155791878700256, + "learning_rate": 4.946551575488776e-05, + "loss": 0.454, + "num_input_tokens_seen": 1634592, + "step": 6080 + }, + { + "epoch": 1.5950196592398427, + "grad_norm": 0.6236544251441956, + "learning_rate": 4.946316084678486e-05, + "loss": 0.4252, + "num_input_tokens_seen": 1635760, + "step": 6085 + }, + { + "epoch": 1.5963302752293578, + "grad_norm": 0.8401573896408081, + "learning_rate": 4.946080081858466e-05, + "loss": 0.6135, + "num_input_tokens_seen": 1637040, + "step": 6090 + }, + { + "epoch": 1.5976408912188729, + "grad_norm": 1.0233005285263062, + "learning_rate": 4.945843567078111e-05, + "loss": 0.4372, + "num_input_tokens_seen": 1638464, + "step": 6095 + }, + { + "epoch": 1.598951507208388, + "grad_norm": 0.631492555141449, + "learning_rate": 4.9456065403869246e-05, + "loss": 0.6158, + "num_input_tokens_seen": 1639792, + "step": 6100 + }, + { + "epoch": 1.600262123197903, + "grad_norm": 0.5715435147285461, + "learning_rate": 4.9453690018345144e-05, + "loss": 0.3245, + "num_input_tokens_seen": 1641056, + "step": 6105 + }, + { + "epoch": 1.601572739187418, + "grad_norm": 1.1965419054031372, + "learning_rate": 4.945130951470597e-05, + "loss": 0.7569, + "num_input_tokens_seen": 1642208, + "step": 6110 + }, + { + "epoch": 1.602883355176933, + "grad_norm": 2.5920658111572266, + "learning_rate": 4.944892389344997e-05, + "loss": 0.3095, + "num_input_tokens_seen": 1643664, + "step": 6115 + }, + { + "epoch": 1.6041939711664481, + "grad_norm": 0.7760657668113708, + "learning_rate": 4.9446533155076444e-05, + "loss": 0.4127, + "num_input_tokens_seen": 1644976, + "step": 6120 + }, + { + "epoch": 1.6055045871559632, + "grad_norm": 1.386507272720337, + "learning_rate": 4.9444137300085766e-05, + "loss": 0.7607, + "num_input_tokens_seen": 1646352, + "step": 6125 + }, + { + "epoch": 1.6068152031454783, + "grad_norm": 1.52906334400177, + "learning_rate": 4.944173632897938e-05, + "loss": 0.4382, + "num_input_tokens_seen": 1647296, + "step": 6130 + }, + { + "epoch": 1.6081258191349934, + "grad_norm": 0.7264705896377563, + "learning_rate": 4.943933024225983e-05, + "loss": 0.6308, + "num_input_tokens_seen": 1649152, + "step": 6135 + }, + { + "epoch": 1.6094364351245085, + "grad_norm": 1.3736597299575806, + "learning_rate": 4.943691904043067e-05, + "loss": 0.5584, + "num_input_tokens_seen": 1650208, + "step": 6140 + }, + { + "epoch": 1.6107470511140236, + "grad_norm": 0.6210423111915588, + "learning_rate": 4.943450272399659e-05, + "loss": 0.5498, + "num_input_tokens_seen": 1650960, + "step": 6145 + }, + { + "epoch": 1.6120576671035387, + "grad_norm": 2.7398641109466553, + "learning_rate": 4.94320812934633e-05, + "loss": 0.4105, + "num_input_tokens_seen": 1652240, + "step": 6150 + }, + { + "epoch": 1.6133682830930538, + "grad_norm": 0.791733980178833, + "learning_rate": 4.942965474933762e-05, + "loss": 0.5949, + "num_input_tokens_seen": 1653568, + "step": 6155 + }, + { + "epoch": 1.614678899082569, + "grad_norm": 0.8892794847488403, + "learning_rate": 4.9427223092127406e-05, + "loss": 0.6382, + "num_input_tokens_seen": 1654944, + "step": 6160 + }, + { + "epoch": 1.615989515072084, + "grad_norm": 0.46745553612709045, + "learning_rate": 4.942478632234161e-05, + "loss": 0.6448, + "num_input_tokens_seen": 1656352, + "step": 6165 + }, + { + "epoch": 1.617300131061599, + "grad_norm": 0.9493942856788635, + "learning_rate": 4.942234444049024e-05, + "loss": 0.4469, + "num_input_tokens_seen": 1657792, + "step": 6170 + }, + { + "epoch": 1.6186107470511142, + "grad_norm": 0.9222638010978699, + "learning_rate": 4.9419897447084375e-05, + "loss": 0.5089, + "num_input_tokens_seen": 1659072, + "step": 6175 + }, + { + "epoch": 1.619921363040629, + "grad_norm": 0.4049319326877594, + "learning_rate": 4.9417445342636166e-05, + "loss": 0.2221, + "num_input_tokens_seen": 1660688, + "step": 6180 + }, + { + "epoch": 1.6212319790301442, + "grad_norm": 1.2048194408416748, + "learning_rate": 4.941498812765884e-05, + "loss": 0.4258, + "num_input_tokens_seen": 1662352, + "step": 6185 + }, + { + "epoch": 1.6225425950196593, + "grad_norm": 4.655052661895752, + "learning_rate": 4.941252580266668e-05, + "loss": 0.7969, + "num_input_tokens_seen": 1663168, + "step": 6190 + }, + { + "epoch": 1.6238532110091743, + "grad_norm": 0.9608544707298279, + "learning_rate": 4.941005836817506e-05, + "loss": 0.5362, + "num_input_tokens_seen": 1664896, + "step": 6195 + }, + { + "epoch": 1.6251638269986892, + "grad_norm": 1.0532746315002441, + "learning_rate": 4.940758582470039e-05, + "loss": 0.4574, + "num_input_tokens_seen": 1666048, + "step": 6200 + }, + { + "epoch": 1.6264744429882043, + "grad_norm": 0.9422631859779358, + "learning_rate": 4.940510817276019e-05, + "loss": 0.6305, + "num_input_tokens_seen": 1667184, + "step": 6205 + }, + { + "epoch": 1.6277850589777194, + "grad_norm": 0.5413415431976318, + "learning_rate": 4.940262541287302e-05, + "loss": 0.2986, + "num_input_tokens_seen": 1668496, + "step": 6210 + }, + { + "epoch": 1.6290956749672345, + "grad_norm": 1.0389409065246582, + "learning_rate": 4.94001375455585e-05, + "loss": 0.8126, + "num_input_tokens_seen": 1669776, + "step": 6215 + }, + { + "epoch": 1.6304062909567496, + "grad_norm": 0.4057476222515106, + "learning_rate": 4.939764457133737e-05, + "loss": 0.5268, + "num_input_tokens_seen": 1671296, + "step": 6220 + }, + { + "epoch": 1.6317169069462647, + "grad_norm": 1.2724474668502808, + "learning_rate": 4.939514649073137e-05, + "loss": 0.4387, + "num_input_tokens_seen": 1672672, + "step": 6225 + }, + { + "epoch": 1.6330275229357798, + "grad_norm": 1.0522066354751587, + "learning_rate": 4.939264330426337e-05, + "loss": 0.4052, + "num_input_tokens_seen": 1673904, + "step": 6230 + }, + { + "epoch": 1.634338138925295, + "grad_norm": 1.1790437698364258, + "learning_rate": 4.9390135012457275e-05, + "loss": 0.4185, + "num_input_tokens_seen": 1675296, + "step": 6235 + }, + { + "epoch": 1.63564875491481, + "grad_norm": 0.7877163290977478, + "learning_rate": 4.938762161583807e-05, + "loss": 0.7141, + "num_input_tokens_seen": 1676400, + "step": 6240 + }, + { + "epoch": 1.636959370904325, + "grad_norm": 0.8405367136001587, + "learning_rate": 4.9385103114931784e-05, + "loss": 0.4396, + "num_input_tokens_seen": 1678064, + "step": 6245 + }, + { + "epoch": 1.6382699868938402, + "grad_norm": 0.8973724246025085, + "learning_rate": 4.9382579510265556e-05, + "loss": 0.5207, + "num_input_tokens_seen": 1679824, + "step": 6250 + }, + { + "epoch": 1.6395806028833553, + "grad_norm": 0.20480619370937347, + "learning_rate": 4.9380050802367573e-05, + "loss": 0.4025, + "num_input_tokens_seen": 1680960, + "step": 6255 + }, + { + "epoch": 1.6408912188728704, + "grad_norm": 0.7966034412384033, + "learning_rate": 4.937751699176708e-05, + "loss": 0.5129, + "num_input_tokens_seen": 1682144, + "step": 6260 + }, + { + "epoch": 1.6422018348623855, + "grad_norm": 0.9965670108795166, + "learning_rate": 4.93749780789944e-05, + "loss": 0.5223, + "num_input_tokens_seen": 1683744, + "step": 6265 + }, + { + "epoch": 1.6435124508519003, + "grad_norm": 0.8693351149559021, + "learning_rate": 4.937243406458093e-05, + "loss": 0.3233, + "num_input_tokens_seen": 1685136, + "step": 6270 + }, + { + "epoch": 1.6448230668414154, + "grad_norm": 5.024900913238525, + "learning_rate": 4.9369884949059115e-05, + "loss": 0.5983, + "num_input_tokens_seen": 1686400, + "step": 6275 + }, + { + "epoch": 1.6461336828309305, + "grad_norm": 1.7022522687911987, + "learning_rate": 4.936733073296249e-05, + "loss": 0.4134, + "num_input_tokens_seen": 1687424, + "step": 6280 + }, + { + "epoch": 1.6474442988204456, + "grad_norm": 1.5939947366714478, + "learning_rate": 4.936477141682565e-05, + "loss": 0.4579, + "num_input_tokens_seen": 1688976, + "step": 6285 + }, + { + "epoch": 1.6487549148099607, + "grad_norm": 0.7218136191368103, + "learning_rate": 4.9362207001184246e-05, + "loss": 0.3171, + "num_input_tokens_seen": 1690608, + "step": 6290 + }, + { + "epoch": 1.6500655307994756, + "grad_norm": 0.8068904280662537, + "learning_rate": 4.9359637486575016e-05, + "loss": 0.5504, + "num_input_tokens_seen": 1691760, + "step": 6295 + }, + { + "epoch": 1.6513761467889907, + "grad_norm": 0.6433715224266052, + "learning_rate": 4.9357062873535746e-05, + "loss": 0.5074, + "num_input_tokens_seen": 1693280, + "step": 6300 + }, + { + "epoch": 1.6526867627785058, + "grad_norm": 0.6422799825668335, + "learning_rate": 4.9354483162605305e-05, + "loss": 0.4299, + "num_input_tokens_seen": 1694272, + "step": 6305 + }, + { + "epoch": 1.6539973787680209, + "grad_norm": 0.6802157163619995, + "learning_rate": 4.9351898354323625e-05, + "loss": 0.4654, + "num_input_tokens_seen": 1695696, + "step": 6310 + }, + { + "epoch": 1.655307994757536, + "grad_norm": 1.3568419218063354, + "learning_rate": 4.934930844923168e-05, + "loss": 0.4604, + "num_input_tokens_seen": 1696704, + "step": 6315 + }, + { + "epoch": 1.656618610747051, + "grad_norm": 1.001653790473938, + "learning_rate": 4.934671344787156e-05, + "loss": 0.4551, + "num_input_tokens_seen": 1697936, + "step": 6320 + }, + { + "epoch": 1.6579292267365662, + "grad_norm": 0.6678786873817444, + "learning_rate": 4.9344113350786386e-05, + "loss": 0.4205, + "num_input_tokens_seen": 1699456, + "step": 6325 + }, + { + "epoch": 1.6592398427260813, + "grad_norm": 1.431483268737793, + "learning_rate": 4.934150815852034e-05, + "loss": 0.4247, + "num_input_tokens_seen": 1700512, + "step": 6330 + }, + { + "epoch": 1.6605504587155964, + "grad_norm": 1.2536334991455078, + "learning_rate": 4.933889787161871e-05, + "loss": 0.4676, + "num_input_tokens_seen": 1701760, + "step": 6335 + }, + { + "epoch": 1.6618610747051115, + "grad_norm": 0.7694019079208374, + "learning_rate": 4.9336282490627803e-05, + "loss": 0.6863, + "num_input_tokens_seen": 1702880, + "step": 6340 + }, + { + "epoch": 1.6631716906946266, + "grad_norm": 1.1319549083709717, + "learning_rate": 4.9333662016095016e-05, + "loss": 0.6484, + "num_input_tokens_seen": 1704064, + "step": 6345 + }, + { + "epoch": 1.6644823066841417, + "grad_norm": 0.6193718314170837, + "learning_rate": 4.9331036448568826e-05, + "loss": 0.4714, + "num_input_tokens_seen": 1705264, + "step": 6350 + }, + { + "epoch": 1.6657929226736568, + "grad_norm": 0.6493822932243347, + "learning_rate": 4.932840578859873e-05, + "loss": 0.3859, + "num_input_tokens_seen": 1706736, + "step": 6355 + }, + { + "epoch": 1.6671035386631718, + "grad_norm": 1.0903420448303223, + "learning_rate": 4.932577003673535e-05, + "loss": 0.4932, + "num_input_tokens_seen": 1708000, + "step": 6360 + }, + { + "epoch": 1.6684141546526867, + "grad_norm": 0.8564603328704834, + "learning_rate": 4.932312919353034e-05, + "loss": 0.4994, + "num_input_tokens_seen": 1709536, + "step": 6365 + }, + { + "epoch": 1.6697247706422018, + "grad_norm": 0.5818183422088623, + "learning_rate": 4.932048325953641e-05, + "loss": 0.4205, + "num_input_tokens_seen": 1710816, + "step": 6370 + }, + { + "epoch": 1.671035386631717, + "grad_norm": 0.8893911838531494, + "learning_rate": 4.931783223530736e-05, + "loss": 0.4272, + "num_input_tokens_seen": 1712288, + "step": 6375 + }, + { + "epoch": 1.672346002621232, + "grad_norm": 0.9081647396087646, + "learning_rate": 4.9315176121398035e-05, + "loss": 0.4461, + "num_input_tokens_seen": 1714192, + "step": 6380 + }, + { + "epoch": 1.673656618610747, + "grad_norm": 0.5056579113006592, + "learning_rate": 4.931251491836436e-05, + "loss": 0.3382, + "num_input_tokens_seen": 1715280, + "step": 6385 + }, + { + "epoch": 1.674967234600262, + "grad_norm": 0.7602084875106812, + "learning_rate": 4.930984862676332e-05, + "loss": 0.3578, + "num_input_tokens_seen": 1716864, + "step": 6390 + }, + { + "epoch": 1.676277850589777, + "grad_norm": 1.0548148155212402, + "learning_rate": 4.930717724715298e-05, + "loss": 0.475, + "num_input_tokens_seen": 1717984, + "step": 6395 + }, + { + "epoch": 1.6775884665792922, + "grad_norm": 0.7905126810073853, + "learning_rate": 4.930450078009243e-05, + "loss": 0.7982, + "num_input_tokens_seen": 1719120, + "step": 6400 + }, + { + "epoch": 1.6788990825688073, + "grad_norm": 1.338127851486206, + "learning_rate": 4.930181922614186e-05, + "loss": 0.7724, + "num_input_tokens_seen": 1721136, + "step": 6405 + }, + { + "epoch": 1.6802096985583224, + "grad_norm": 0.5895509719848633, + "learning_rate": 4.929913258586252e-05, + "loss": 0.5078, + "num_input_tokens_seen": 1722528, + "step": 6410 + }, + { + "epoch": 1.6815203145478375, + "grad_norm": 0.5212574005126953, + "learning_rate": 4.9296440859816715e-05, + "loss": 0.4908, + "num_input_tokens_seen": 1723856, + "step": 6415 + }, + { + "epoch": 1.6828309305373526, + "grad_norm": 0.9958983063697815, + "learning_rate": 4.9293744048567825e-05, + "loss": 0.4685, + "num_input_tokens_seen": 1724976, + "step": 6420 + }, + { + "epoch": 1.6841415465268676, + "grad_norm": 1.3557188510894775, + "learning_rate": 4.929104215268027e-05, + "loss": 0.7041, + "num_input_tokens_seen": 1726656, + "step": 6425 + }, + { + "epoch": 1.6854521625163827, + "grad_norm": 0.7305877804756165, + "learning_rate": 4.9288335172719575e-05, + "loss": 0.6325, + "num_input_tokens_seen": 1727920, + "step": 6430 + }, + { + "epoch": 1.6867627785058978, + "grad_norm": 0.833158016204834, + "learning_rate": 4.9285623109252286e-05, + "loss": 0.5742, + "num_input_tokens_seen": 1729200, + "step": 6435 + }, + { + "epoch": 1.688073394495413, + "grad_norm": 0.7541870474815369, + "learning_rate": 4.928290596284606e-05, + "loss": 0.2873, + "num_input_tokens_seen": 1730352, + "step": 6440 + }, + { + "epoch": 1.689384010484928, + "grad_norm": 0.522338330745697, + "learning_rate": 4.928018373406955e-05, + "loss": 0.3992, + "num_input_tokens_seen": 1731616, + "step": 6445 + }, + { + "epoch": 1.6906946264744431, + "grad_norm": 3.1217801570892334, + "learning_rate": 4.927745642349255e-05, + "loss": 0.6935, + "num_input_tokens_seen": 1732960, + "step": 6450 + }, + { + "epoch": 1.6920052424639582, + "grad_norm": 1.103107213973999, + "learning_rate": 4.927472403168587e-05, + "loss": 0.6478, + "num_input_tokens_seen": 1734160, + "step": 6455 + }, + { + "epoch": 1.693315858453473, + "grad_norm": 0.8987711668014526, + "learning_rate": 4.92719865592214e-05, + "loss": 0.5031, + "num_input_tokens_seen": 1735392, + "step": 6460 + }, + { + "epoch": 1.6946264744429882, + "grad_norm": 0.8488577604293823, + "learning_rate": 4.9269244006672065e-05, + "loss": 0.5273, + "num_input_tokens_seen": 1736576, + "step": 6465 + }, + { + "epoch": 1.6959370904325033, + "grad_norm": 3.5305163860321045, + "learning_rate": 4.9266496374611904e-05, + "loss": 0.5046, + "num_input_tokens_seen": 1737792, + "step": 6470 + }, + { + "epoch": 1.6972477064220184, + "grad_norm": 0.5556241869926453, + "learning_rate": 4.926374366361598e-05, + "loss": 0.5997, + "num_input_tokens_seen": 1739184, + "step": 6475 + }, + { + "epoch": 1.6985583224115333, + "grad_norm": 0.711281955242157, + "learning_rate": 4.926098587426043e-05, + "loss": 0.4816, + "num_input_tokens_seen": 1740624, + "step": 6480 + }, + { + "epoch": 1.6998689384010484, + "grad_norm": 0.6675758957862854, + "learning_rate": 4.9258223007122454e-05, + "loss": 0.3772, + "num_input_tokens_seen": 1742480, + "step": 6485 + }, + { + "epoch": 1.7011795543905635, + "grad_norm": 0.417436420917511, + "learning_rate": 4.925545506278032e-05, + "loss": 0.4265, + "num_input_tokens_seen": 1743808, + "step": 6490 + }, + { + "epoch": 1.7024901703800785, + "grad_norm": 0.6566947102546692, + "learning_rate": 4.9252682041813356e-05, + "loss": 0.6889, + "num_input_tokens_seen": 1744704, + "step": 6495 + }, + { + "epoch": 1.7038007863695936, + "grad_norm": 0.6343250274658203, + "learning_rate": 4.924990394480195e-05, + "loss": 0.3576, + "num_input_tokens_seen": 1745872, + "step": 6500 + }, + { + "epoch": 1.7051114023591087, + "grad_norm": 0.580471396446228, + "learning_rate": 4.9247120772327534e-05, + "loss": 0.4876, + "num_input_tokens_seen": 1747184, + "step": 6505 + }, + { + "epoch": 1.7064220183486238, + "grad_norm": 1.0677541494369507, + "learning_rate": 4.924433252497264e-05, + "loss": 0.6374, + "num_input_tokens_seen": 1748880, + "step": 6510 + }, + { + "epoch": 1.707732634338139, + "grad_norm": 2.0175139904022217, + "learning_rate": 4.924153920332084e-05, + "loss": 0.687, + "num_input_tokens_seen": 1750400, + "step": 6515 + }, + { + "epoch": 1.709043250327654, + "grad_norm": 0.7512770891189575, + "learning_rate": 4.923874080795677e-05, + "loss": 0.3968, + "num_input_tokens_seen": 1752000, + "step": 6520 + }, + { + "epoch": 1.7103538663171691, + "grad_norm": 0.8875799775123596, + "learning_rate": 4.923593733946614e-05, + "loss": 0.6991, + "num_input_tokens_seen": 1753520, + "step": 6525 + }, + { + "epoch": 1.7116644823066842, + "grad_norm": 1.3329126834869385, + "learning_rate": 4.92331287984357e-05, + "loss": 0.5648, + "num_input_tokens_seen": 1754704, + "step": 6530 + }, + { + "epoch": 1.7129750982961993, + "grad_norm": 1.1651734113693237, + "learning_rate": 4.9230315185453266e-05, + "loss": 0.4611, + "num_input_tokens_seen": 1755488, + "step": 6535 + }, + { + "epoch": 1.7142857142857144, + "grad_norm": 0.7663881778717041, + "learning_rate": 4.922749650110773e-05, + "loss": 0.3577, + "num_input_tokens_seen": 1756688, + "step": 6540 + }, + { + "epoch": 1.7155963302752295, + "grad_norm": 0.6030871272087097, + "learning_rate": 4.9224672745989045e-05, + "loss": 0.5254, + "num_input_tokens_seen": 1758256, + "step": 6545 + }, + { + "epoch": 1.7169069462647444, + "grad_norm": 3.2221078872680664, + "learning_rate": 4.922184392068821e-05, + "loss": 0.6037, + "num_input_tokens_seen": 1759392, + "step": 6550 + }, + { + "epoch": 1.7182175622542595, + "grad_norm": 1.7344849109649658, + "learning_rate": 4.921901002579729e-05, + "loss": 0.4874, + "num_input_tokens_seen": 1760800, + "step": 6555 + }, + { + "epoch": 1.7195281782437746, + "grad_norm": 1.0160547494888306, + "learning_rate": 4.921617106190942e-05, + "loss": 0.4038, + "num_input_tokens_seen": 1761872, + "step": 6560 + }, + { + "epoch": 1.7208387942332897, + "grad_norm": 2.3625481128692627, + "learning_rate": 4.9213327029618796e-05, + "loss": 0.5099, + "num_input_tokens_seen": 1763392, + "step": 6565 + }, + { + "epoch": 1.7221494102228048, + "grad_norm": 0.6659313440322876, + "learning_rate": 4.921047792952066e-05, + "loss": 0.4802, + "num_input_tokens_seen": 1764832, + "step": 6570 + }, + { + "epoch": 1.7234600262123196, + "grad_norm": 0.6815107464790344, + "learning_rate": 4.920762376221132e-05, + "loss": 0.3665, + "num_input_tokens_seen": 1766304, + "step": 6575 + }, + { + "epoch": 1.7247706422018347, + "grad_norm": 0.5442764759063721, + "learning_rate": 4.920476452828815e-05, + "loss": 0.3015, + "num_input_tokens_seen": 1767632, + "step": 6580 + }, + { + "epoch": 1.7260812581913498, + "grad_norm": 1.2885605096817017, + "learning_rate": 4.9201900228349594e-05, + "loss": 0.3602, + "num_input_tokens_seen": 1768944, + "step": 6585 + }, + { + "epoch": 1.727391874180865, + "grad_norm": 1.053438425064087, + "learning_rate": 4.9199030862995144e-05, + "loss": 0.5524, + "num_input_tokens_seen": 1770672, + "step": 6590 + }, + { + "epoch": 1.72870249017038, + "grad_norm": 1.0912518501281738, + "learning_rate": 4.919615643282533e-05, + "loss": 0.5582, + "num_input_tokens_seen": 1771808, + "step": 6595 + }, + { + "epoch": 1.7300131061598951, + "grad_norm": 0.840757429599762, + "learning_rate": 4.91932769384418e-05, + "loss": 0.5623, + "num_input_tokens_seen": 1773056, + "step": 6600 + }, + { + "epoch": 1.7313237221494102, + "grad_norm": 0.7794971466064453, + "learning_rate": 4.91903923804472e-05, + "loss": 0.6349, + "num_input_tokens_seen": 1774528, + "step": 6605 + }, + { + "epoch": 1.7326343381389253, + "grad_norm": 0.9980721473693848, + "learning_rate": 4.918750275944528e-05, + "loss": 0.7762, + "num_input_tokens_seen": 1775680, + "step": 6610 + }, + { + "epoch": 1.7339449541284404, + "grad_norm": 0.558678925037384, + "learning_rate": 4.9184608076040816e-05, + "loss": 0.4057, + "num_input_tokens_seen": 1776944, + "step": 6615 + }, + { + "epoch": 1.7352555701179555, + "grad_norm": 0.526607871055603, + "learning_rate": 4.918170833083967e-05, + "loss": 0.4373, + "num_input_tokens_seen": 1778432, + "step": 6620 + }, + { + "epoch": 1.7365661861074706, + "grad_norm": 0.42114555835723877, + "learning_rate": 4.9178803524448756e-05, + "loss": 0.5407, + "num_input_tokens_seen": 1779520, + "step": 6625 + }, + { + "epoch": 1.7378768020969857, + "grad_norm": 0.9509638547897339, + "learning_rate": 4.9175893657476046e-05, + "loss": 0.7205, + "num_input_tokens_seen": 1780864, + "step": 6630 + }, + { + "epoch": 1.7391874180865008, + "grad_norm": 0.8715909123420715, + "learning_rate": 4.917297873053056e-05, + "loss": 0.4595, + "num_input_tokens_seen": 1782608, + "step": 6635 + }, + { + "epoch": 1.7404980340760159, + "grad_norm": 3.7936999797821045, + "learning_rate": 4.91700587442224e-05, + "loss": 0.4745, + "num_input_tokens_seen": 1783680, + "step": 6640 + }, + { + "epoch": 1.7418086500655308, + "grad_norm": 5.210254669189453, + "learning_rate": 4.91671336991627e-05, + "loss": 0.5299, + "num_input_tokens_seen": 1784720, + "step": 6645 + }, + { + "epoch": 1.7431192660550459, + "grad_norm": 2.140658140182495, + "learning_rate": 4.916420359596368e-05, + "loss": 0.6534, + "num_input_tokens_seen": 1785792, + "step": 6650 + }, + { + "epoch": 1.744429882044561, + "grad_norm": 0.2978002727031708, + "learning_rate": 4.91612684352386e-05, + "loss": 0.3575, + "num_input_tokens_seen": 1787008, + "step": 6655 + }, + { + "epoch": 1.745740498034076, + "grad_norm": 0.8013126254081726, + "learning_rate": 4.915832821760178e-05, + "loss": 0.3958, + "num_input_tokens_seen": 1788384, + "step": 6660 + }, + { + "epoch": 1.747051114023591, + "grad_norm": 0.7392286062240601, + "learning_rate": 4.915538294366861e-05, + "loss": 0.7378, + "num_input_tokens_seen": 1789888, + "step": 6665 + }, + { + "epoch": 1.748361730013106, + "grad_norm": 0.7653586864471436, + "learning_rate": 4.915243261405553e-05, + "loss": 0.4678, + "num_input_tokens_seen": 1791088, + "step": 6670 + }, + { + "epoch": 1.7496723460026211, + "grad_norm": 1.0873972177505493, + "learning_rate": 4.9149477229380023e-05, + "loss": 0.4892, + "num_input_tokens_seen": 1792384, + "step": 6675 + }, + { + "epoch": 1.7509829619921362, + "grad_norm": 1.0188963413238525, + "learning_rate": 4.914651679026068e-05, + "loss": 0.5519, + "num_input_tokens_seen": 1793776, + "step": 6680 + }, + { + "epoch": 1.7522935779816513, + "grad_norm": 0.584422767162323, + "learning_rate": 4.9143551297317076e-05, + "loss": 0.3658, + "num_input_tokens_seen": 1794880, + "step": 6685 + }, + { + "epoch": 1.7536041939711664, + "grad_norm": 0.571021556854248, + "learning_rate": 4.9140580751169906e-05, + "loss": 0.3853, + "num_input_tokens_seen": 1796016, + "step": 6690 + }, + { + "epoch": 1.7549148099606815, + "grad_norm": 0.8382250666618347, + "learning_rate": 4.91376051524409e-05, + "loss": 0.5429, + "num_input_tokens_seen": 1797264, + "step": 6695 + }, + { + "epoch": 1.7562254259501966, + "grad_norm": 0.4347458481788635, + "learning_rate": 4.913462450175285e-05, + "loss": 0.36, + "num_input_tokens_seen": 1798368, + "step": 6700 + }, + { + "epoch": 1.7575360419397117, + "grad_norm": 0.45614391565322876, + "learning_rate": 4.913163879972957e-05, + "loss": 0.3876, + "num_input_tokens_seen": 1799744, + "step": 6705 + }, + { + "epoch": 1.7588466579292268, + "grad_norm": 1.325192928314209, + "learning_rate": 4.9128648046996004e-05, + "loss": 0.4997, + "num_input_tokens_seen": 1800992, + "step": 6710 + }, + { + "epoch": 1.7601572739187419, + "grad_norm": 0.8860205411911011, + "learning_rate": 4.912565224417809e-05, + "loss": 0.5297, + "num_input_tokens_seen": 1802640, + "step": 6715 + }, + { + "epoch": 1.761467889908257, + "grad_norm": 0.6347373127937317, + "learning_rate": 4.9122651391902855e-05, + "loss": 0.3772, + "num_input_tokens_seen": 1803920, + "step": 6720 + }, + { + "epoch": 1.762778505897772, + "grad_norm": 0.8989322781562805, + "learning_rate": 4.9119645490798347e-05, + "loss": 0.5314, + "num_input_tokens_seen": 1805088, + "step": 6725 + }, + { + "epoch": 1.7640891218872872, + "grad_norm": 0.5123246312141418, + "learning_rate": 4.911663454149372e-05, + "loss": 0.6178, + "num_input_tokens_seen": 1806512, + "step": 6730 + }, + { + "epoch": 1.765399737876802, + "grad_norm": 0.7749911546707153, + "learning_rate": 4.911361854461916e-05, + "loss": 0.6562, + "num_input_tokens_seen": 1808064, + "step": 6735 + }, + { + "epoch": 1.7667103538663171, + "grad_norm": 0.49338027834892273, + "learning_rate": 4.91105975008059e-05, + "loss": 0.2593, + "num_input_tokens_seen": 1809216, + "step": 6740 + }, + { + "epoch": 1.7680209698558322, + "grad_norm": 1.5555344820022583, + "learning_rate": 4.910757141068624e-05, + "loss": 0.472, + "num_input_tokens_seen": 1810368, + "step": 6745 + }, + { + "epoch": 1.7693315858453473, + "grad_norm": 1.5924657583236694, + "learning_rate": 4.910454027489354e-05, + "loss": 0.3699, + "num_input_tokens_seen": 1812000, + "step": 6750 + }, + { + "epoch": 1.7706422018348624, + "grad_norm": 0.8857523798942566, + "learning_rate": 4.910150409406221e-05, + "loss": 0.5355, + "num_input_tokens_seen": 1813232, + "step": 6755 + }, + { + "epoch": 1.7719528178243773, + "grad_norm": 2.973491907119751, + "learning_rate": 4.909846286882772e-05, + "loss": 0.4123, + "num_input_tokens_seen": 1814832, + "step": 6760 + }, + { + "epoch": 1.7732634338138924, + "grad_norm": 1.179169774055481, + "learning_rate": 4.909541659982659e-05, + "loss": 0.3869, + "num_input_tokens_seen": 1815808, + "step": 6765 + }, + { + "epoch": 1.7745740498034075, + "grad_norm": 0.8357797861099243, + "learning_rate": 4.9092365287696396e-05, + "loss": 0.3981, + "num_input_tokens_seen": 1817296, + "step": 6770 + }, + { + "epoch": 1.7758846657929226, + "grad_norm": 1.1456998586654663, + "learning_rate": 4.908930893307578e-05, + "loss": 0.3364, + "num_input_tokens_seen": 1818528, + "step": 6775 + }, + { + "epoch": 1.7771952817824377, + "grad_norm": 1.2488480806350708, + "learning_rate": 4.9086247536604425e-05, + "loss": 0.4895, + "num_input_tokens_seen": 1819888, + "step": 6780 + }, + { + "epoch": 1.7785058977719528, + "grad_norm": 0.6851770877838135, + "learning_rate": 4.908318109892308e-05, + "loss": 0.4091, + "num_input_tokens_seen": 1821232, + "step": 6785 + }, + { + "epoch": 1.7798165137614679, + "grad_norm": 0.7501692771911621, + "learning_rate": 4.9080109620673546e-05, + "loss": 0.4059, + "num_input_tokens_seen": 1822544, + "step": 6790 + }, + { + "epoch": 1.781127129750983, + "grad_norm": 0.9417160749435425, + "learning_rate": 4.9077033102498666e-05, + "loss": 0.5584, + "num_input_tokens_seen": 1823712, + "step": 6795 + }, + { + "epoch": 1.782437745740498, + "grad_norm": 1.199450135231018, + "learning_rate": 4.907395154504236e-05, + "loss": 0.4401, + "num_input_tokens_seen": 1824736, + "step": 6800 + }, + { + "epoch": 1.7837483617300132, + "grad_norm": 0.7497369050979614, + "learning_rate": 4.9070864948949604e-05, + "loss": 0.3558, + "num_input_tokens_seen": 1825920, + "step": 6805 + }, + { + "epoch": 1.7850589777195283, + "grad_norm": 0.48894500732421875, + "learning_rate": 4.9067773314866396e-05, + "loss": 0.5049, + "num_input_tokens_seen": 1827104, + "step": 6810 + }, + { + "epoch": 1.7863695937090434, + "grad_norm": 1.1222546100616455, + "learning_rate": 4.906467664343982e-05, + "loss": 0.3654, + "num_input_tokens_seen": 1829024, + "step": 6815 + }, + { + "epoch": 1.7876802096985585, + "grad_norm": 1.4793237447738647, + "learning_rate": 4.9061574935318006e-05, + "loss": 0.3996, + "num_input_tokens_seen": 1830336, + "step": 6820 + }, + { + "epoch": 1.7889908256880735, + "grad_norm": 1.2153472900390625, + "learning_rate": 4.905846819115013e-05, + "loss": 0.4375, + "num_input_tokens_seen": 1831360, + "step": 6825 + }, + { + "epoch": 1.7903014416775884, + "grad_norm": 0.6897327303886414, + "learning_rate": 4.905535641158644e-05, + "loss": 0.235, + "num_input_tokens_seen": 1832800, + "step": 6830 + }, + { + "epoch": 1.7916120576671035, + "grad_norm": 1.8877778053283691, + "learning_rate": 4.9052239597278206e-05, + "loss": 0.3785, + "num_input_tokens_seen": 1833792, + "step": 6835 + }, + { + "epoch": 1.7929226736566186, + "grad_norm": 1.6364970207214355, + "learning_rate": 4.904911774887779e-05, + "loss": 0.4301, + "num_input_tokens_seen": 1835008, + "step": 6840 + }, + { + "epoch": 1.7942332896461337, + "grad_norm": 0.7497086524963379, + "learning_rate": 4.904599086703858e-05, + "loss": 0.4316, + "num_input_tokens_seen": 1836256, + "step": 6845 + }, + { + "epoch": 1.7955439056356488, + "grad_norm": 1.7718979120254517, + "learning_rate": 4.904285895241503e-05, + "loss": 0.4564, + "num_input_tokens_seen": 1837408, + "step": 6850 + }, + { + "epoch": 1.7968545216251637, + "grad_norm": 1.1004889011383057, + "learning_rate": 4.903972200566265e-05, + "loss": 0.5461, + "num_input_tokens_seen": 1838736, + "step": 6855 + }, + { + "epoch": 1.7981651376146788, + "grad_norm": 0.6233354806900024, + "learning_rate": 4.903658002743798e-05, + "loss": 0.4436, + "num_input_tokens_seen": 1840016, + "step": 6860 + }, + { + "epoch": 1.7994757536041939, + "grad_norm": 0.6322982907295227, + "learning_rate": 4.9033433018398654e-05, + "loss": 0.4894, + "num_input_tokens_seen": 1841040, + "step": 6865 + }, + { + "epoch": 1.800786369593709, + "grad_norm": 0.34289878606796265, + "learning_rate": 4.9030280979203314e-05, + "loss": 0.3675, + "num_input_tokens_seen": 1842128, + "step": 6870 + }, + { + "epoch": 1.802096985583224, + "grad_norm": 0.9684337973594666, + "learning_rate": 4.90271239105117e-05, + "loss": 0.2298, + "num_input_tokens_seen": 1843632, + "step": 6875 + }, + { + "epoch": 1.8034076015727392, + "grad_norm": 1.826442003250122, + "learning_rate": 4.902396181298455e-05, + "loss": 0.4957, + "num_input_tokens_seen": 1844704, + "step": 6880 + }, + { + "epoch": 1.8047182175622543, + "grad_norm": 0.4571308195590973, + "learning_rate": 4.902079468728371e-05, + "loss": 0.5148, + "num_input_tokens_seen": 1846080, + "step": 6885 + }, + { + "epoch": 1.8060288335517694, + "grad_norm": 1.0962010622024536, + "learning_rate": 4.901762253407205e-05, + "loss": 0.4484, + "num_input_tokens_seen": 1847024, + "step": 6890 + }, + { + "epoch": 1.8073394495412844, + "grad_norm": 1.021587610244751, + "learning_rate": 4.9014445354013496e-05, + "loss": 0.4711, + "num_input_tokens_seen": 1848352, + "step": 6895 + }, + { + "epoch": 1.8086500655307995, + "grad_norm": 1.035240650177002, + "learning_rate": 4.901126314777301e-05, + "loss": 0.5996, + "num_input_tokens_seen": 1849360, + "step": 6900 + }, + { + "epoch": 1.8099606815203146, + "grad_norm": 0.42725569009780884, + "learning_rate": 4.9008075916016646e-05, + "loss": 0.4712, + "num_input_tokens_seen": 1850848, + "step": 6905 + }, + { + "epoch": 1.8112712975098297, + "grad_norm": 0.7864119410514832, + "learning_rate": 4.900488365941147e-05, + "loss": 0.4763, + "num_input_tokens_seen": 1851824, + "step": 6910 + }, + { + "epoch": 1.8125819134993448, + "grad_norm": 0.539478063583374, + "learning_rate": 4.900168637862563e-05, + "loss": 0.4406, + "num_input_tokens_seen": 1853056, + "step": 6915 + }, + { + "epoch": 1.81389252948886, + "grad_norm": 2.0310111045837402, + "learning_rate": 4.899848407432829e-05, + "loss": 0.4498, + "num_input_tokens_seen": 1854480, + "step": 6920 + }, + { + "epoch": 1.8152031454783748, + "grad_norm": 0.670201301574707, + "learning_rate": 4.8995276747189714e-05, + "loss": 0.4228, + "num_input_tokens_seen": 1856384, + "step": 6925 + }, + { + "epoch": 1.81651376146789, + "grad_norm": 0.3974205255508423, + "learning_rate": 4.899206439788117e-05, + "loss": 0.2889, + "num_input_tokens_seen": 1857584, + "step": 6930 + }, + { + "epoch": 1.817824377457405, + "grad_norm": 1.2865532636642456, + "learning_rate": 4.898884702707499e-05, + "loss": 0.7707, + "num_input_tokens_seen": 1858960, + "step": 6935 + }, + { + "epoch": 1.81913499344692, + "grad_norm": 0.9740791916847229, + "learning_rate": 4.898562463544459e-05, + "loss": 0.5675, + "num_input_tokens_seen": 1859920, + "step": 6940 + }, + { + "epoch": 1.820445609436435, + "grad_norm": 1.0461416244506836, + "learning_rate": 4.89823972236644e-05, + "loss": 0.5555, + "num_input_tokens_seen": 1861280, + "step": 6945 + }, + { + "epoch": 1.82175622542595, + "grad_norm": 1.0717275142669678, + "learning_rate": 4.89791647924099e-05, + "loss": 0.584, + "num_input_tokens_seen": 1862576, + "step": 6950 + }, + { + "epoch": 1.8230668414154652, + "grad_norm": 0.408332884311676, + "learning_rate": 4.8975927342357646e-05, + "loss": 0.5641, + "num_input_tokens_seen": 1864192, + "step": 6955 + }, + { + "epoch": 1.8243774574049803, + "grad_norm": 0.7200020551681519, + "learning_rate": 4.897268487418523e-05, + "loss": 0.4247, + "num_input_tokens_seen": 1865664, + "step": 6960 + }, + { + "epoch": 1.8256880733944953, + "grad_norm": 5.510375022888184, + "learning_rate": 4.896943738857128e-05, + "loss": 0.5206, + "num_input_tokens_seen": 1867072, + "step": 6965 + }, + { + "epoch": 1.8269986893840104, + "grad_norm": 0.9251489639282227, + "learning_rate": 4.896618488619551e-05, + "loss": 0.4323, + "num_input_tokens_seen": 1868160, + "step": 6970 + }, + { + "epoch": 1.8283093053735255, + "grad_norm": 1.3190200328826904, + "learning_rate": 4.8962927367738655e-05, + "loss": 0.4551, + "num_input_tokens_seen": 1869104, + "step": 6975 + }, + { + "epoch": 1.8296199213630406, + "grad_norm": 0.7021294832229614, + "learning_rate": 4.89596648338825e-05, + "loss": 0.4155, + "num_input_tokens_seen": 1870448, + "step": 6980 + }, + { + "epoch": 1.8309305373525557, + "grad_norm": 0.7777636647224426, + "learning_rate": 4.89563972853099e-05, + "loss": 0.6055, + "num_input_tokens_seen": 1871904, + "step": 6985 + }, + { + "epoch": 1.8322411533420708, + "grad_norm": 0.6127062439918518, + "learning_rate": 4.895312472270475e-05, + "loss": 0.3008, + "num_input_tokens_seen": 1873216, + "step": 6990 + }, + { + "epoch": 1.833551769331586, + "grad_norm": 0.5836616158485413, + "learning_rate": 4.894984714675197e-05, + "loss": 0.5842, + "num_input_tokens_seen": 1874480, + "step": 6995 + }, + { + "epoch": 1.834862385321101, + "grad_norm": 0.48729240894317627, + "learning_rate": 4.8946564558137566e-05, + "loss": 0.5178, + "num_input_tokens_seen": 1876064, + "step": 7000 + }, + { + "epoch": 1.8361730013106161, + "grad_norm": 1.0347524881362915, + "learning_rate": 4.894327695754858e-05, + "loss": 0.2944, + "num_input_tokens_seen": 1877488, + "step": 7005 + }, + { + "epoch": 1.8374836173001312, + "grad_norm": 1.142155408859253, + "learning_rate": 4.893998434567309e-05, + "loss": 0.4502, + "num_input_tokens_seen": 1878880, + "step": 7010 + }, + { + "epoch": 1.838794233289646, + "grad_norm": 1.3073495626449585, + "learning_rate": 4.893668672320025e-05, + "loss": 0.441, + "num_input_tokens_seen": 1880416, + "step": 7015 + }, + { + "epoch": 1.8401048492791612, + "grad_norm": 0.5884913206100464, + "learning_rate": 4.893338409082023e-05, + "loss": 0.4927, + "num_input_tokens_seen": 1881936, + "step": 7020 + }, + { + "epoch": 1.8414154652686763, + "grad_norm": 1.204938292503357, + "learning_rate": 4.8930076449224285e-05, + "loss": 0.4503, + "num_input_tokens_seen": 1883056, + "step": 7025 + }, + { + "epoch": 1.8427260812581914, + "grad_norm": 0.7785066366195679, + "learning_rate": 4.892676379910468e-05, + "loss": 0.4519, + "num_input_tokens_seen": 1884192, + "step": 7030 + }, + { + "epoch": 1.8440366972477065, + "grad_norm": 0.9470977783203125, + "learning_rate": 4.8923446141154746e-05, + "loss": 0.4078, + "num_input_tokens_seen": 1885968, + "step": 7035 + }, + { + "epoch": 1.8453473132372213, + "grad_norm": 1.026617169380188, + "learning_rate": 4.8920123476068894e-05, + "loss": 0.5244, + "num_input_tokens_seen": 1887024, + "step": 7040 + }, + { + "epoch": 1.8466579292267364, + "grad_norm": 0.9073201417922974, + "learning_rate": 4.891679580454251e-05, + "loss": 0.5197, + "num_input_tokens_seen": 1888288, + "step": 7045 + }, + { + "epoch": 1.8479685452162515, + "grad_norm": 0.8884222507476807, + "learning_rate": 4.89134631272721e-05, + "loss": 0.5468, + "num_input_tokens_seen": 1890128, + "step": 7050 + }, + { + "epoch": 1.8492791612057666, + "grad_norm": 0.533350944519043, + "learning_rate": 4.891012544495517e-05, + "loss": 0.5916, + "num_input_tokens_seen": 1891584, + "step": 7055 + }, + { + "epoch": 1.8505897771952817, + "grad_norm": 1.316293716430664, + "learning_rate": 4.890678275829029e-05, + "loss": 0.4005, + "num_input_tokens_seen": 1892944, + "step": 7060 + }, + { + "epoch": 1.8519003931847968, + "grad_norm": 0.9386557936668396, + "learning_rate": 4.89034350679771e-05, + "loss": 0.4141, + "num_input_tokens_seen": 1894144, + "step": 7065 + }, + { + "epoch": 1.853211009174312, + "grad_norm": 1.145486831665039, + "learning_rate": 4.890008237471625e-05, + "loss": 0.665, + "num_input_tokens_seen": 1895056, + "step": 7070 + }, + { + "epoch": 1.854521625163827, + "grad_norm": 0.7477907538414001, + "learning_rate": 4.889672467920945e-05, + "loss": 0.47, + "num_input_tokens_seen": 1896416, + "step": 7075 + }, + { + "epoch": 1.855832241153342, + "grad_norm": 0.9682533740997314, + "learning_rate": 4.889336198215947e-05, + "loss": 0.5784, + "num_input_tokens_seen": 1897456, + "step": 7080 + }, + { + "epoch": 1.8571428571428572, + "grad_norm": 0.5230990648269653, + "learning_rate": 4.888999428427011e-05, + "loss": 0.4671, + "num_input_tokens_seen": 1898816, + "step": 7085 + }, + { + "epoch": 1.8584534731323723, + "grad_norm": 0.9047384858131409, + "learning_rate": 4.8886621586246226e-05, + "loss": 0.5511, + "num_input_tokens_seen": 1900176, + "step": 7090 + }, + { + "epoch": 1.8597640891218874, + "grad_norm": 0.5483137369155884, + "learning_rate": 4.8883243888793715e-05, + "loss": 0.3163, + "num_input_tokens_seen": 1901424, + "step": 7095 + }, + { + "epoch": 1.8610747051114025, + "grad_norm": 0.8166833519935608, + "learning_rate": 4.887986119261953e-05, + "loss": 0.3249, + "num_input_tokens_seen": 1903616, + "step": 7100 + }, + { + "epoch": 1.8623853211009176, + "grad_norm": 0.6874544620513916, + "learning_rate": 4.887647349843165e-05, + "loss": 0.3334, + "num_input_tokens_seen": 1904928, + "step": 7105 + }, + { + "epoch": 1.8636959370904325, + "grad_norm": 1.3692233562469482, + "learning_rate": 4.887308080693913e-05, + "loss": 0.4514, + "num_input_tokens_seen": 1907632, + "step": 7110 + }, + { + "epoch": 1.8650065530799476, + "grad_norm": 0.8647543787956238, + "learning_rate": 4.886968311885204e-05, + "loss": 0.3851, + "num_input_tokens_seen": 1908672, + "step": 7115 + }, + { + "epoch": 1.8663171690694627, + "grad_norm": 0.5334885120391846, + "learning_rate": 4.8866280434881516e-05, + "loss": 1.016, + "num_input_tokens_seen": 1909808, + "step": 7120 + }, + { + "epoch": 1.8676277850589778, + "grad_norm": 0.8628392815589905, + "learning_rate": 4.886287275573974e-05, + "loss": 0.4676, + "num_input_tokens_seen": 1911248, + "step": 7125 + }, + { + "epoch": 1.8689384010484926, + "grad_norm": 1.2907053232192993, + "learning_rate": 4.885946008213993e-05, + "loss": 0.5536, + "num_input_tokens_seen": 1912688, + "step": 7130 + }, + { + "epoch": 1.8702490170380077, + "grad_norm": 0.918567419052124, + "learning_rate": 4.885604241479635e-05, + "loss": 0.5481, + "num_input_tokens_seen": 1914384, + "step": 7135 + }, + { + "epoch": 1.8715596330275228, + "grad_norm": 0.8056138157844543, + "learning_rate": 4.885261975442431e-05, + "loss": 0.3867, + "num_input_tokens_seen": 1915696, + "step": 7140 + }, + { + "epoch": 1.872870249017038, + "grad_norm": 0.6426905989646912, + "learning_rate": 4.884919210174017e-05, + "loss": 0.6435, + "num_input_tokens_seen": 1917424, + "step": 7145 + }, + { + "epoch": 1.874180865006553, + "grad_norm": 0.708587110042572, + "learning_rate": 4.8845759457461326e-05, + "loss": 0.3624, + "num_input_tokens_seen": 1918800, + "step": 7150 + }, + { + "epoch": 1.875491480996068, + "grad_norm": 0.7014760375022888, + "learning_rate": 4.8842321822306235e-05, + "loss": 0.3654, + "num_input_tokens_seen": 1920240, + "step": 7155 + }, + { + "epoch": 1.8768020969855832, + "grad_norm": 0.5397488474845886, + "learning_rate": 4.8838879196994385e-05, + "loss": 0.5913, + "num_input_tokens_seen": 1921696, + "step": 7160 + }, + { + "epoch": 1.8781127129750983, + "grad_norm": 1.3475213050842285, + "learning_rate": 4.88354315822463e-05, + "loss": 0.6107, + "num_input_tokens_seen": 1922848, + "step": 7165 + }, + { + "epoch": 1.8794233289646134, + "grad_norm": 0.6694055795669556, + "learning_rate": 4.8831978978783585e-05, + "loss": 0.4324, + "num_input_tokens_seen": 1924160, + "step": 7170 + }, + { + "epoch": 1.8807339449541285, + "grad_norm": 0.458158016204834, + "learning_rate": 4.882852138732884e-05, + "loss": 0.4357, + "num_input_tokens_seen": 1925792, + "step": 7175 + }, + { + "epoch": 1.8820445609436436, + "grad_norm": 0.5510730743408203, + "learning_rate": 4.882505880860574e-05, + "loss": 0.3262, + "num_input_tokens_seen": 1927312, + "step": 7180 + }, + { + "epoch": 1.8833551769331587, + "grad_norm": 1.3121135234832764, + "learning_rate": 4.8821591243339e-05, + "loss": 0.4171, + "num_input_tokens_seen": 1929248, + "step": 7185 + }, + { + "epoch": 1.8846657929226738, + "grad_norm": 0.7931621074676514, + "learning_rate": 4.8818118692254375e-05, + "loss": 0.3429, + "num_input_tokens_seen": 1930464, + "step": 7190 + }, + { + "epoch": 1.8859764089121889, + "grad_norm": 0.8120977282524109, + "learning_rate": 4.881464115607865e-05, + "loss": 0.488, + "num_input_tokens_seen": 1931552, + "step": 7195 + }, + { + "epoch": 1.8872870249017037, + "grad_norm": 0.32810381054878235, + "learning_rate": 4.88111586355397e-05, + "loss": 0.2718, + "num_input_tokens_seen": 1932768, + "step": 7200 + }, + { + "epoch": 1.8885976408912188, + "grad_norm": 1.0873271226882935, + "learning_rate": 4.880767113136637e-05, + "loss": 0.3948, + "num_input_tokens_seen": 1934064, + "step": 7205 + }, + { + "epoch": 1.889908256880734, + "grad_norm": 0.6928123831748962, + "learning_rate": 4.880417864428862e-05, + "loss": 0.4868, + "num_input_tokens_seen": 1935744, + "step": 7210 + }, + { + "epoch": 1.891218872870249, + "grad_norm": 0.6251839995384216, + "learning_rate": 4.880068117503741e-05, + "loss": 0.5078, + "num_input_tokens_seen": 1937232, + "step": 7215 + }, + { + "epoch": 1.8925294888597641, + "grad_norm": 1.1661803722381592, + "learning_rate": 4.879717872434475e-05, + "loss": 0.5124, + "num_input_tokens_seen": 1938704, + "step": 7220 + }, + { + "epoch": 1.893840104849279, + "grad_norm": 0.9877117872238159, + "learning_rate": 4.879367129294371e-05, + "loss": 0.6806, + "num_input_tokens_seen": 1939696, + "step": 7225 + }, + { + "epoch": 1.895150720838794, + "grad_norm": 0.6034039855003357, + "learning_rate": 4.879015888156837e-05, + "loss": 0.5547, + "num_input_tokens_seen": 1940960, + "step": 7230 + }, + { + "epoch": 1.8964613368283092, + "grad_norm": 1.9992045164108276, + "learning_rate": 4.8786641490953886e-05, + "loss": 0.5603, + "num_input_tokens_seen": 1942160, + "step": 7235 + }, + { + "epoch": 1.8977719528178243, + "grad_norm": 0.8058087825775146, + "learning_rate": 4.878311912183644e-05, + "loss": 0.4578, + "num_input_tokens_seen": 1943840, + "step": 7240 + }, + { + "epoch": 1.8990825688073394, + "grad_norm": 0.5121321082115173, + "learning_rate": 4.877959177495326e-05, + "loss": 0.3062, + "num_input_tokens_seen": 1944896, + "step": 7245 + }, + { + "epoch": 1.9003931847968545, + "grad_norm": 0.40254077315330505, + "learning_rate": 4.87760594510426e-05, + "loss": 0.5118, + "num_input_tokens_seen": 1945952, + "step": 7250 + }, + { + "epoch": 1.9017038007863696, + "grad_norm": 1.4367589950561523, + "learning_rate": 4.877252215084379e-05, + "loss": 0.445, + "num_input_tokens_seen": 1947072, + "step": 7255 + }, + { + "epoch": 1.9030144167758847, + "grad_norm": 0.44876226782798767, + "learning_rate": 4.876897987509716e-05, + "loss": 0.3714, + "num_input_tokens_seen": 1948128, + "step": 7260 + }, + { + "epoch": 1.9043250327653998, + "grad_norm": 0.7340888381004333, + "learning_rate": 4.876543262454412e-05, + "loss": 0.498, + "num_input_tokens_seen": 1949344, + "step": 7265 + }, + { + "epoch": 1.9056356487549149, + "grad_norm": 0.7354043126106262, + "learning_rate": 4.8761880399927084e-05, + "loss": 0.4699, + "num_input_tokens_seen": 1950848, + "step": 7270 + }, + { + "epoch": 1.90694626474443, + "grad_norm": 0.953542947769165, + "learning_rate": 4.875832320198954e-05, + "loss": 0.4883, + "num_input_tokens_seen": 1951904, + "step": 7275 + }, + { + "epoch": 1.908256880733945, + "grad_norm": 0.6801841855049133, + "learning_rate": 4.8754761031476005e-05, + "loss": 0.565, + "num_input_tokens_seen": 1953664, + "step": 7280 + }, + { + "epoch": 1.9095674967234602, + "grad_norm": 1.7210577726364136, + "learning_rate": 4.875119388913203e-05, + "loss": 0.4738, + "num_input_tokens_seen": 1955328, + "step": 7285 + }, + { + "epoch": 1.9108781127129753, + "grad_norm": 0.8741795420646667, + "learning_rate": 4.874762177570422e-05, + "loss": 0.4558, + "num_input_tokens_seen": 1956624, + "step": 7290 + }, + { + "epoch": 1.9121887287024901, + "grad_norm": 0.8836877942085266, + "learning_rate": 4.874404469194019e-05, + "loss": 0.4743, + "num_input_tokens_seen": 1957872, + "step": 7295 + }, + { + "epoch": 1.9134993446920052, + "grad_norm": 0.7070772647857666, + "learning_rate": 4.8740462638588644e-05, + "loss": 0.3227, + "num_input_tokens_seen": 1958864, + "step": 7300 + }, + { + "epoch": 1.9148099606815203, + "grad_norm": 0.6176527738571167, + "learning_rate": 4.873687561639929e-05, + "loss": 0.5353, + "num_input_tokens_seen": 1960160, + "step": 7305 + }, + { + "epoch": 1.9161205766710354, + "grad_norm": 1.4306105375289917, + "learning_rate": 4.8733283626122875e-05, + "loss": 0.6231, + "num_input_tokens_seen": 1961248, + "step": 7310 + }, + { + "epoch": 1.9174311926605505, + "grad_norm": 6.827646732330322, + "learning_rate": 4.872968666851121e-05, + "loss": 0.3777, + "num_input_tokens_seen": 1962400, + "step": 7315 + }, + { + "epoch": 1.9187418086500654, + "grad_norm": 0.785407543182373, + "learning_rate": 4.872608474431712e-05, + "loss": 0.546, + "num_input_tokens_seen": 1964352, + "step": 7320 + }, + { + "epoch": 1.9200524246395805, + "grad_norm": 0.7316048741340637, + "learning_rate": 4.87224778542945e-05, + "loss": 0.3009, + "num_input_tokens_seen": 1965376, + "step": 7325 + }, + { + "epoch": 1.9213630406290956, + "grad_norm": 0.6267983317375183, + "learning_rate": 4.871886599919825e-05, + "loss": 0.4398, + "num_input_tokens_seen": 1966512, + "step": 7330 + }, + { + "epoch": 1.9226736566186107, + "grad_norm": 1.0209563970565796, + "learning_rate": 4.871524917978433e-05, + "loss": 0.5583, + "num_input_tokens_seen": 1968112, + "step": 7335 + }, + { + "epoch": 1.9239842726081258, + "grad_norm": 0.3610915243625641, + "learning_rate": 4.8711627396809725e-05, + "loss": 0.3906, + "num_input_tokens_seen": 1969472, + "step": 7340 + }, + { + "epoch": 1.9252948885976409, + "grad_norm": 1.8197987079620361, + "learning_rate": 4.870800065103248e-05, + "loss": 0.5405, + "num_input_tokens_seen": 1970768, + "step": 7345 + }, + { + "epoch": 1.926605504587156, + "grad_norm": 1.8255631923675537, + "learning_rate": 4.870436894321167e-05, + "loss": 0.6332, + "num_input_tokens_seen": 1972432, + "step": 7350 + }, + { + "epoch": 1.927916120576671, + "grad_norm": 4.043508529663086, + "learning_rate": 4.87007322741074e-05, + "loss": 0.7686, + "num_input_tokens_seen": 1974128, + "step": 7355 + }, + { + "epoch": 1.9292267365661862, + "grad_norm": 0.8881657719612122, + "learning_rate": 4.8697090644480804e-05, + "loss": 0.3668, + "num_input_tokens_seen": 1975616, + "step": 7360 + }, + { + "epoch": 1.9305373525557012, + "grad_norm": 0.5266183018684387, + "learning_rate": 4.86934440550941e-05, + "loss": 0.6144, + "num_input_tokens_seen": 1976944, + "step": 7365 + }, + { + "epoch": 1.9318479685452163, + "grad_norm": 2.926176071166992, + "learning_rate": 4.868979250671047e-05, + "loss": 0.5629, + "num_input_tokens_seen": 1977856, + "step": 7370 + }, + { + "epoch": 1.9331585845347314, + "grad_norm": 1.2583023309707642, + "learning_rate": 4.868613600009422e-05, + "loss": 0.4331, + "num_input_tokens_seen": 1979184, + "step": 7375 + }, + { + "epoch": 1.9344692005242465, + "grad_norm": 0.5081851482391357, + "learning_rate": 4.8682474536010624e-05, + "loss": 0.4027, + "num_input_tokens_seen": 1980512, + "step": 7380 + }, + { + "epoch": 1.9357798165137616, + "grad_norm": 0.49581214785575867, + "learning_rate": 4.8678808115226026e-05, + "loss": 0.3142, + "num_input_tokens_seen": 1981936, + "step": 7385 + }, + { + "epoch": 1.9370904325032765, + "grad_norm": 0.7819440960884094, + "learning_rate": 4.867513673850781e-05, + "loss": 0.4944, + "num_input_tokens_seen": 1983376, + "step": 7390 + }, + { + "epoch": 1.9384010484927916, + "grad_norm": 0.39858531951904297, + "learning_rate": 4.867146040662438e-05, + "loss": 0.578, + "num_input_tokens_seen": 1984912, + "step": 7395 + }, + { + "epoch": 1.9397116644823067, + "grad_norm": 1.6749838590621948, + "learning_rate": 4.8667779120345185e-05, + "loss": 0.4446, + "num_input_tokens_seen": 1986112, + "step": 7400 + }, + { + "epoch": 1.9410222804718218, + "grad_norm": 0.5419154763221741, + "learning_rate": 4.866409288044072e-05, + "loss": 0.4738, + "num_input_tokens_seen": 1987824, + "step": 7405 + }, + { + "epoch": 1.9423328964613367, + "grad_norm": 0.5872724056243896, + "learning_rate": 4.86604016876825e-05, + "loss": 0.4426, + "num_input_tokens_seen": 1988752, + "step": 7410 + }, + { + "epoch": 1.9436435124508518, + "grad_norm": 1.1120980978012085, + "learning_rate": 4.8656705542843085e-05, + "loss": 0.5628, + "num_input_tokens_seen": 1990384, + "step": 7415 + }, + { + "epoch": 1.9449541284403669, + "grad_norm": 0.5443400144577026, + "learning_rate": 4.865300444669608e-05, + "loss": 0.6938, + "num_input_tokens_seen": 1991504, + "step": 7420 + }, + { + "epoch": 1.946264744429882, + "grad_norm": 0.7319643497467041, + "learning_rate": 4.864929840001612e-05, + "loss": 0.3184, + "num_input_tokens_seen": 1993040, + "step": 7425 + }, + { + "epoch": 1.947575360419397, + "grad_norm": 0.5409528613090515, + "learning_rate": 4.8645587403578855e-05, + "loss": 0.3466, + "num_input_tokens_seen": 1994176, + "step": 7430 + }, + { + "epoch": 1.9488859764089121, + "grad_norm": 0.6119168400764465, + "learning_rate": 4.864187145816101e-05, + "loss": 0.3685, + "num_input_tokens_seen": 1995232, + "step": 7435 + }, + { + "epoch": 1.9501965923984272, + "grad_norm": 4.066282272338867, + "learning_rate": 4.863815056454032e-05, + "loss": 0.6535, + "num_input_tokens_seen": 1996464, + "step": 7440 + }, + { + "epoch": 1.9515072083879423, + "grad_norm": 0.40754324197769165, + "learning_rate": 4.863442472349555e-05, + "loss": 0.3803, + "num_input_tokens_seen": 1997488, + "step": 7445 + }, + { + "epoch": 1.9528178243774574, + "grad_norm": 0.719669759273529, + "learning_rate": 4.8630693935806536e-05, + "loss": 0.4319, + "num_input_tokens_seen": 1998608, + "step": 7450 + }, + { + "epoch": 1.9541284403669725, + "grad_norm": 1.5384632349014282, + "learning_rate": 4.8626958202254106e-05, + "loss": 0.6473, + "num_input_tokens_seen": 1999728, + "step": 7455 + }, + { + "epoch": 1.9554390563564876, + "grad_norm": 0.5009105205535889, + "learning_rate": 4.862321752362015e-05, + "loss": 0.5921, + "num_input_tokens_seen": 2000848, + "step": 7460 + }, + { + "epoch": 1.9567496723460027, + "grad_norm": 1.4857091903686523, + "learning_rate": 4.8619471900687583e-05, + "loss": 0.7304, + "num_input_tokens_seen": 2001840, + "step": 7465 + }, + { + "epoch": 1.9580602883355178, + "grad_norm": 3.5542824268341064, + "learning_rate": 4.861572133424036e-05, + "loss": 0.4053, + "num_input_tokens_seen": 2003072, + "step": 7470 + }, + { + "epoch": 1.959370904325033, + "grad_norm": 0.5491653680801392, + "learning_rate": 4.8611965825063466e-05, + "loss": 0.4924, + "num_input_tokens_seen": 2004448, + "step": 7475 + }, + { + "epoch": 1.9606815203145478, + "grad_norm": 0.7633111476898193, + "learning_rate": 4.8608205373942915e-05, + "loss": 0.4546, + "num_input_tokens_seen": 2005472, + "step": 7480 + }, + { + "epoch": 1.9619921363040629, + "grad_norm": 0.624883770942688, + "learning_rate": 4.860443998166578e-05, + "loss": 0.3472, + "num_input_tokens_seen": 2006752, + "step": 7485 + }, + { + "epoch": 1.963302752293578, + "grad_norm": 0.9499841928482056, + "learning_rate": 4.860066964902014e-05, + "loss": 0.5099, + "num_input_tokens_seen": 2008048, + "step": 7490 + }, + { + "epoch": 1.964613368283093, + "grad_norm": 0.3850749433040619, + "learning_rate": 4.859689437679512e-05, + "loss": 0.5158, + "num_input_tokens_seen": 2009520, + "step": 7495 + }, + { + "epoch": 1.9659239842726082, + "grad_norm": 1.46077299118042, + "learning_rate": 4.859311416578088e-05, + "loss": 0.4353, + "num_input_tokens_seen": 2010848, + "step": 7500 + }, + { + "epoch": 1.967234600262123, + "grad_norm": 0.46472135186195374, + "learning_rate": 4.8589329016768615e-05, + "loss": 0.7076, + "num_input_tokens_seen": 2012336, + "step": 7505 + }, + { + "epoch": 1.9685452162516381, + "grad_norm": 0.5832013487815857, + "learning_rate": 4.8585538930550547e-05, + "loss": 0.5284, + "num_input_tokens_seen": 2013872, + "step": 7510 + }, + { + "epoch": 1.9698558322411532, + "grad_norm": 1.369580626487732, + "learning_rate": 4.858174390791993e-05, + "loss": 0.3982, + "num_input_tokens_seen": 2015040, + "step": 7515 + }, + { + "epoch": 1.9711664482306683, + "grad_norm": 0.536482036113739, + "learning_rate": 4.857794394967106e-05, + "loss": 0.3339, + "num_input_tokens_seen": 2016256, + "step": 7520 + }, + { + "epoch": 1.9724770642201834, + "grad_norm": 0.4631718695163727, + "learning_rate": 4.8574139056599256e-05, + "loss": 0.3635, + "num_input_tokens_seen": 2017984, + "step": 7525 + }, + { + "epoch": 1.9737876802096985, + "grad_norm": 0.8280507922172546, + "learning_rate": 4.8570329229500884e-05, + "loss": 0.5411, + "num_input_tokens_seen": 2019488, + "step": 7530 + }, + { + "epoch": 1.9750982961992136, + "grad_norm": 1.8173562288284302, + "learning_rate": 4.8566514469173326e-05, + "loss": 0.3798, + "num_input_tokens_seen": 2020736, + "step": 7535 + }, + { + "epoch": 1.9764089121887287, + "grad_norm": 0.30293288826942444, + "learning_rate": 4.8562694776415016e-05, + "loss": 0.4181, + "num_input_tokens_seen": 2022240, + "step": 7540 + }, + { + "epoch": 1.9777195281782438, + "grad_norm": 4.931527614593506, + "learning_rate": 4.8558870152025404e-05, + "loss": 0.7083, + "num_input_tokens_seen": 2023936, + "step": 7545 + }, + { + "epoch": 1.979030144167759, + "grad_norm": 0.5971086025238037, + "learning_rate": 4.855504059680497e-05, + "loss": 0.3999, + "num_input_tokens_seen": 2025248, + "step": 7550 + }, + { + "epoch": 1.980340760157274, + "grad_norm": 0.4442010819911957, + "learning_rate": 4.855120611155524e-05, + "loss": 0.3243, + "num_input_tokens_seen": 2026960, + "step": 7555 + }, + { + "epoch": 1.981651376146789, + "grad_norm": 0.656076967716217, + "learning_rate": 4.854736669707876e-05, + "loss": 0.4281, + "num_input_tokens_seen": 2028384, + "step": 7560 + }, + { + "epoch": 1.9829619921363042, + "grad_norm": 0.9518714547157288, + "learning_rate": 4.854352235417913e-05, + "loss": 0.521, + "num_input_tokens_seen": 2029856, + "step": 7565 + }, + { + "epoch": 1.9842726081258193, + "grad_norm": 1.1004638671875, + "learning_rate": 4.853967308366094e-05, + "loss": 0.3568, + "num_input_tokens_seen": 2031104, + "step": 7570 + }, + { + "epoch": 1.9855832241153342, + "grad_norm": 1.0777003765106201, + "learning_rate": 4.8535818886329855e-05, + "loss": 0.5399, + "num_input_tokens_seen": 2032368, + "step": 7575 + }, + { + "epoch": 1.9868938401048493, + "grad_norm": 0.554032564163208, + "learning_rate": 4.853195976299254e-05, + "loss": 0.4129, + "num_input_tokens_seen": 2033536, + "step": 7580 + }, + { + "epoch": 1.9882044560943644, + "grad_norm": 0.6029179692268372, + "learning_rate": 4.85280957144567e-05, + "loss": 0.4641, + "num_input_tokens_seen": 2035600, + "step": 7585 + }, + { + "epoch": 1.9895150720838795, + "grad_norm": 0.8031386733055115, + "learning_rate": 4.852422674153109e-05, + "loss": 0.5591, + "num_input_tokens_seen": 2036928, + "step": 7590 + }, + { + "epoch": 1.9908256880733946, + "grad_norm": 0.9503940939903259, + "learning_rate": 4.852035284502547e-05, + "loss": 0.3377, + "num_input_tokens_seen": 2038144, + "step": 7595 + }, + { + "epoch": 1.9921363040629094, + "grad_norm": 0.5242370367050171, + "learning_rate": 4.851647402575063e-05, + "loss": 0.5126, + "num_input_tokens_seen": 2039328, + "step": 7600 + }, + { + "epoch": 1.9934469200524245, + "grad_norm": 0.6537632346153259, + "learning_rate": 4.851259028451842e-05, + "loss": 0.5164, + "num_input_tokens_seen": 2040496, + "step": 7605 + }, + { + "epoch": 1.9947575360419396, + "grad_norm": 0.7467095851898193, + "learning_rate": 4.850870162214168e-05, + "loss": 0.4688, + "num_input_tokens_seen": 2041904, + "step": 7610 + }, + { + "epoch": 1.9960681520314547, + "grad_norm": 0.19932733476161957, + "learning_rate": 4.850480803943432e-05, + "loss": 0.6058, + "num_input_tokens_seen": 2043376, + "step": 7615 + }, + { + "epoch": 1.9973787680209698, + "grad_norm": 1.1845849752426147, + "learning_rate": 4.8500909537211246e-05, + "loss": 0.4699, + "num_input_tokens_seen": 2044736, + "step": 7620 + }, + { + "epoch": 1.998689384010485, + "grad_norm": 0.47170931100845337, + "learning_rate": 4.849700611628841e-05, + "loss": 0.345, + "num_input_tokens_seen": 2046016, + "step": 7625 + }, + { + "epoch": 2.0, + "grad_norm": 0.6497707366943359, + "learning_rate": 4.84930977774828e-05, + "loss": 0.5013, + "num_input_tokens_seen": 2047128, + "step": 7630 + }, + { + "epoch": 2.000524246395806, + "eval_loss": 0.4960760176181793, + "eval_runtime": 16.9908, + "eval_samples_per_second": 49.909, + "eval_steps_per_second": 24.955, + "num_input_tokens_seen": 2047464, + "step": 7632 + }, + { + "epoch": 2.001310615989515, + "grad_norm": 0.4593764841556549, + "learning_rate": 4.8489184521612416e-05, + "loss": 0.4151, + "num_input_tokens_seen": 2048456, + "step": 7635 + }, + { + "epoch": 2.00262123197903, + "grad_norm": 1.5028430223464966, + "learning_rate": 4.848526634949629e-05, + "loss": 0.4935, + "num_input_tokens_seen": 2049368, + "step": 7640 + }, + { + "epoch": 2.0039318479685453, + "grad_norm": 1.1087982654571533, + "learning_rate": 4.848134326195451e-05, + "loss": 0.3583, + "num_input_tokens_seen": 2050728, + "step": 7645 + }, + { + "epoch": 2.0052424639580604, + "grad_norm": 0.9120528101921082, + "learning_rate": 4.847741525980815e-05, + "loss": 0.3566, + "num_input_tokens_seen": 2051864, + "step": 7650 + }, + { + "epoch": 2.0065530799475755, + "grad_norm": 0.6256507635116577, + "learning_rate": 4.8473482343879344e-05, + "loss": 0.6319, + "num_input_tokens_seen": 2053240, + "step": 7655 + }, + { + "epoch": 2.0078636959370906, + "grad_norm": 1.0096039772033691, + "learning_rate": 4.8469544514991235e-05, + "loss": 0.4577, + "num_input_tokens_seen": 2054280, + "step": 7660 + }, + { + "epoch": 2.0091743119266057, + "grad_norm": 0.7076054215431213, + "learning_rate": 4.846560177396802e-05, + "loss": 0.3051, + "num_input_tokens_seen": 2055400, + "step": 7665 + }, + { + "epoch": 2.0104849279161208, + "grad_norm": 0.5003446936607361, + "learning_rate": 4.8461654121634905e-05, + "loss": 0.4865, + "num_input_tokens_seen": 2056952, + "step": 7670 + }, + { + "epoch": 2.011795543905636, + "grad_norm": 0.9251864552497864, + "learning_rate": 4.8457701558818106e-05, + "loss": 0.381, + "num_input_tokens_seen": 2058104, + "step": 7675 + }, + { + "epoch": 2.0131061598951505, + "grad_norm": 0.7078182697296143, + "learning_rate": 4.845374408634491e-05, + "loss": 0.349, + "num_input_tokens_seen": 2059224, + "step": 7680 + }, + { + "epoch": 2.0144167758846656, + "grad_norm": 0.886078953742981, + "learning_rate": 4.84497817050436e-05, + "loss": 0.4083, + "num_input_tokens_seen": 2060776, + "step": 7685 + }, + { + "epoch": 2.0157273918741807, + "grad_norm": 2.4640026092529297, + "learning_rate": 4.8445814415743504e-05, + "loss": 0.5723, + "num_input_tokens_seen": 2062120, + "step": 7690 + }, + { + "epoch": 2.017038007863696, + "grad_norm": 0.6498444080352783, + "learning_rate": 4.844184221927495e-05, + "loss": 0.3991, + "num_input_tokens_seen": 2063208, + "step": 7695 + }, + { + "epoch": 2.018348623853211, + "grad_norm": 0.5295084118843079, + "learning_rate": 4.843786511646933e-05, + "loss": 0.6443, + "num_input_tokens_seen": 2065096, + "step": 7700 + }, + { + "epoch": 2.019659239842726, + "grad_norm": 1.14736008644104, + "learning_rate": 4.843388310815904e-05, + "loss": 0.5346, + "num_input_tokens_seen": 2067304, + "step": 7705 + }, + { + "epoch": 2.020969855832241, + "grad_norm": 0.5590280294418335, + "learning_rate": 4.8429896195177504e-05, + "loss": 0.4206, + "num_input_tokens_seen": 2068728, + "step": 7710 + }, + { + "epoch": 2.022280471821756, + "grad_norm": 0.9256020784378052, + "learning_rate": 4.842590437835918e-05, + "loss": 0.51, + "num_input_tokens_seen": 2069880, + "step": 7715 + }, + { + "epoch": 2.0235910878112713, + "grad_norm": 0.9304442405700684, + "learning_rate": 4.8421907658539545e-05, + "loss": 0.3809, + "num_input_tokens_seen": 2071096, + "step": 7720 + }, + { + "epoch": 2.0249017038007864, + "grad_norm": 0.7289294004440308, + "learning_rate": 4.84179060365551e-05, + "loss": 0.4317, + "num_input_tokens_seen": 2074312, + "step": 7725 + }, + { + "epoch": 2.0262123197903015, + "grad_norm": 3.8770077228546143, + "learning_rate": 4.84138995132434e-05, + "loss": 0.5679, + "num_input_tokens_seen": 2075688, + "step": 7730 + }, + { + "epoch": 2.0275229357798166, + "grad_norm": 0.5064663887023926, + "learning_rate": 4.840988808944298e-05, + "loss": 0.3894, + "num_input_tokens_seen": 2076808, + "step": 7735 + }, + { + "epoch": 2.0288335517693317, + "grad_norm": 0.36864325404167175, + "learning_rate": 4.8405871765993433e-05, + "loss": 0.3782, + "num_input_tokens_seen": 2078440, + "step": 7740 + }, + { + "epoch": 2.0301441677588468, + "grad_norm": 0.4060420095920563, + "learning_rate": 4.840185054373536e-05, + "loss": 0.4853, + "num_input_tokens_seen": 2079656, + "step": 7745 + }, + { + "epoch": 2.031454783748362, + "grad_norm": 2.7165348529815674, + "learning_rate": 4.839782442351042e-05, + "loss": 0.5816, + "num_input_tokens_seen": 2080712, + "step": 7750 + }, + { + "epoch": 2.032765399737877, + "grad_norm": 0.726050078868866, + "learning_rate": 4.839379340616124e-05, + "loss": 0.4526, + "num_input_tokens_seen": 2082088, + "step": 7755 + }, + { + "epoch": 2.034076015727392, + "grad_norm": 0.854279637336731, + "learning_rate": 4.838975749253153e-05, + "loss": 0.4667, + "num_input_tokens_seen": 2083208, + "step": 7760 + }, + { + "epoch": 2.035386631716907, + "grad_norm": 1.2696704864501953, + "learning_rate": 4.838571668346599e-05, + "loss": 0.42, + "num_input_tokens_seen": 2084312, + "step": 7765 + }, + { + "epoch": 2.036697247706422, + "grad_norm": 0.4197635054588318, + "learning_rate": 4.838167097981036e-05, + "loss": 0.3791, + "num_input_tokens_seen": 2085880, + "step": 7770 + }, + { + "epoch": 2.038007863695937, + "grad_norm": 0.5984938144683838, + "learning_rate": 4.8377620382411385e-05, + "loss": 0.4417, + "num_input_tokens_seen": 2087080, + "step": 7775 + }, + { + "epoch": 2.039318479685452, + "grad_norm": 0.8948092460632324, + "learning_rate": 4.837356489211686e-05, + "loss": 0.4089, + "num_input_tokens_seen": 2088440, + "step": 7780 + }, + { + "epoch": 2.040629095674967, + "grad_norm": 0.8302451372146606, + "learning_rate": 4.836950450977559e-05, + "loss": 0.412, + "num_input_tokens_seen": 2089432, + "step": 7785 + }, + { + "epoch": 2.041939711664482, + "grad_norm": 0.6056162714958191, + "learning_rate": 4.836543923623741e-05, + "loss": 0.3061, + "num_input_tokens_seen": 2090952, + "step": 7790 + }, + { + "epoch": 2.0432503276539973, + "grad_norm": 0.7902677059173584, + "learning_rate": 4.8361369072353164e-05, + "loss": 0.6361, + "num_input_tokens_seen": 2092648, + "step": 7795 + }, + { + "epoch": 2.0445609436435124, + "grad_norm": 0.9948599338531494, + "learning_rate": 4.835729401897474e-05, + "loss": 0.4726, + "num_input_tokens_seen": 2093896, + "step": 7800 + }, + { + "epoch": 2.0458715596330275, + "grad_norm": 0.8530715703964233, + "learning_rate": 4.835321407695503e-05, + "loss": 0.4295, + "num_input_tokens_seen": 2095320, + "step": 7805 + }, + { + "epoch": 2.0471821756225426, + "grad_norm": 0.6154961585998535, + "learning_rate": 4.834912924714797e-05, + "loss": 0.3812, + "num_input_tokens_seen": 2096488, + "step": 7810 + }, + { + "epoch": 2.0484927916120577, + "grad_norm": 1.2235221862792969, + "learning_rate": 4.834503953040851e-05, + "loss": 0.5354, + "num_input_tokens_seen": 2097880, + "step": 7815 + }, + { + "epoch": 2.0498034076015728, + "grad_norm": 1.82542085647583, + "learning_rate": 4.83409449275926e-05, + "loss": 0.4753, + "num_input_tokens_seen": 2099160, + "step": 7820 + }, + { + "epoch": 2.051114023591088, + "grad_norm": 1.6447728872299194, + "learning_rate": 4.8336845439557256e-05, + "loss": 0.3719, + "num_input_tokens_seen": 2100632, + "step": 7825 + }, + { + "epoch": 2.052424639580603, + "grad_norm": 1.0368828773498535, + "learning_rate": 4.833274106716048e-05, + "loss": 0.2624, + "num_input_tokens_seen": 2101688, + "step": 7830 + }, + { + "epoch": 2.053735255570118, + "grad_norm": 0.43253815174102783, + "learning_rate": 4.832863181126132e-05, + "loss": 0.367, + "num_input_tokens_seen": 2103176, + "step": 7835 + }, + { + "epoch": 2.055045871559633, + "grad_norm": 0.5514971017837524, + "learning_rate": 4.832451767271983e-05, + "loss": 0.7646, + "num_input_tokens_seen": 2104360, + "step": 7840 + }, + { + "epoch": 2.0563564875491482, + "grad_norm": 2.573939085006714, + "learning_rate": 4.832039865239708e-05, + "loss": 0.3736, + "num_input_tokens_seen": 2106072, + "step": 7845 + }, + { + "epoch": 2.0576671035386633, + "grad_norm": 0.6173258423805237, + "learning_rate": 4.83162747511552e-05, + "loss": 0.3456, + "num_input_tokens_seen": 2107288, + "step": 7850 + }, + { + "epoch": 2.0589777195281784, + "grad_norm": 1.2035576105117798, + "learning_rate": 4.83121459698573e-05, + "loss": 0.4113, + "num_input_tokens_seen": 2108456, + "step": 7855 + }, + { + "epoch": 2.0602883355176935, + "grad_norm": 0.8480732440948486, + "learning_rate": 4.830801230936752e-05, + "loss": 0.5566, + "num_input_tokens_seen": 2109896, + "step": 7860 + }, + { + "epoch": 2.061598951507208, + "grad_norm": 1.4803491830825806, + "learning_rate": 4.830387377055105e-05, + "loss": 0.4262, + "num_input_tokens_seen": 2111560, + "step": 7865 + }, + { + "epoch": 2.0629095674967233, + "grad_norm": 1.6836637258529663, + "learning_rate": 4.829973035427406e-05, + "loss": 0.5241, + "num_input_tokens_seen": 2112584, + "step": 7870 + }, + { + "epoch": 2.0642201834862384, + "grad_norm": 0.885237991809845, + "learning_rate": 4.8295582061403766e-05, + "loss": 0.331, + "num_input_tokens_seen": 2114424, + "step": 7875 + }, + { + "epoch": 2.0655307994757535, + "grad_norm": 0.6910213828086853, + "learning_rate": 4.8291428892808396e-05, + "loss": 0.4728, + "num_input_tokens_seen": 2115672, + "step": 7880 + }, + { + "epoch": 2.0668414154652686, + "grad_norm": 0.6041919589042664, + "learning_rate": 4.8287270849357205e-05, + "loss": 0.5759, + "num_input_tokens_seen": 2117336, + "step": 7885 + }, + { + "epoch": 2.0681520314547837, + "grad_norm": 0.5077120661735535, + "learning_rate": 4.828310793192046e-05, + "loss": 0.3356, + "num_input_tokens_seen": 2118792, + "step": 7890 + }, + { + "epoch": 2.0694626474442988, + "grad_norm": 1.0061262845993042, + "learning_rate": 4.827894014136945e-05, + "loss": 0.4875, + "num_input_tokens_seen": 2120072, + "step": 7895 + }, + { + "epoch": 2.070773263433814, + "grad_norm": 0.6687341332435608, + "learning_rate": 4.827476747857651e-05, + "loss": 0.3931, + "num_input_tokens_seen": 2120952, + "step": 7900 + }, + { + "epoch": 2.072083879423329, + "grad_norm": 1.231694221496582, + "learning_rate": 4.827058994441494e-05, + "loss": 0.3772, + "num_input_tokens_seen": 2121880, + "step": 7905 + }, + { + "epoch": 2.073394495412844, + "grad_norm": 0.6956042647361755, + "learning_rate": 4.82664075397591e-05, + "loss": 0.5332, + "num_input_tokens_seen": 2123128, + "step": 7910 + }, + { + "epoch": 2.074705111402359, + "grad_norm": 0.6110199093818665, + "learning_rate": 4.826222026548437e-05, + "loss": 0.4384, + "num_input_tokens_seen": 2125016, + "step": 7915 + }, + { + "epoch": 2.0760157273918742, + "grad_norm": 2.4904558658599854, + "learning_rate": 4.825802812246713e-05, + "loss": 1.2038, + "num_input_tokens_seen": 2126088, + "step": 7920 + }, + { + "epoch": 2.0773263433813893, + "grad_norm": 0.6725988984107971, + "learning_rate": 4.8253831111584794e-05, + "loss": 0.3906, + "num_input_tokens_seen": 2127576, + "step": 7925 + }, + { + "epoch": 2.0786369593709044, + "grad_norm": 0.6088208556175232, + "learning_rate": 4.8249629233715776e-05, + "loss": 0.3431, + "num_input_tokens_seen": 2128776, + "step": 7930 + }, + { + "epoch": 2.0799475753604195, + "grad_norm": 0.9950403571128845, + "learning_rate": 4.824542248973954e-05, + "loss": 0.4383, + "num_input_tokens_seen": 2130056, + "step": 7935 + }, + { + "epoch": 2.0812581913499346, + "grad_norm": 0.5317594408988953, + "learning_rate": 4.824121088053654e-05, + "loss": 0.4033, + "num_input_tokens_seen": 2131432, + "step": 7940 + }, + { + "epoch": 2.0825688073394497, + "grad_norm": 1.2510358095169067, + "learning_rate": 4.8236994406988255e-05, + "loss": 0.3579, + "num_input_tokens_seen": 2132584, + "step": 7945 + }, + { + "epoch": 2.083879423328965, + "grad_norm": 0.5859067440032959, + "learning_rate": 4.82327730699772e-05, + "loss": 0.3566, + "num_input_tokens_seen": 2134184, + "step": 7950 + }, + { + "epoch": 2.08519003931848, + "grad_norm": 0.8575617671012878, + "learning_rate": 4.822854687038688e-05, + "loss": 0.4886, + "num_input_tokens_seen": 2135624, + "step": 7955 + }, + { + "epoch": 2.0865006553079946, + "grad_norm": 0.910754919052124, + "learning_rate": 4.8224315809101836e-05, + "loss": 0.4466, + "num_input_tokens_seen": 2136712, + "step": 7960 + }, + { + "epoch": 2.0878112712975097, + "grad_norm": 0.5199431777000427, + "learning_rate": 4.822007988700762e-05, + "loss": 0.6535, + "num_input_tokens_seen": 2138264, + "step": 7965 + }, + { + "epoch": 2.0891218872870247, + "grad_norm": 1.294746994972229, + "learning_rate": 4.821583910499081e-05, + "loss": 0.4829, + "num_input_tokens_seen": 2139368, + "step": 7970 + }, + { + "epoch": 2.09043250327654, + "grad_norm": 2.084693670272827, + "learning_rate": 4.8211593463938985e-05, + "loss": 0.6046, + "num_input_tokens_seen": 2140456, + "step": 7975 + }, + { + "epoch": 2.091743119266055, + "grad_norm": 0.5595287680625916, + "learning_rate": 4.8207342964740755e-05, + "loss": 0.4958, + "num_input_tokens_seen": 2142024, + "step": 7980 + }, + { + "epoch": 2.09305373525557, + "grad_norm": 0.8445576429367065, + "learning_rate": 4.8203087608285744e-05, + "loss": 0.5429, + "num_input_tokens_seen": 2143208, + "step": 7985 + }, + { + "epoch": 2.094364351245085, + "grad_norm": 0.5256744623184204, + "learning_rate": 4.819882739546458e-05, + "loss": 0.2736, + "num_input_tokens_seen": 2144712, + "step": 7990 + }, + { + "epoch": 2.0956749672346002, + "grad_norm": 0.5515844821929932, + "learning_rate": 4.8194562327168946e-05, + "loss": 0.4935, + "num_input_tokens_seen": 2146392, + "step": 7995 + }, + { + "epoch": 2.0969855832241153, + "grad_norm": 0.5259466767311096, + "learning_rate": 4.819029240429148e-05, + "loss": 0.5112, + "num_input_tokens_seen": 2148136, + "step": 8000 + }, + { + "epoch": 2.0982961992136304, + "grad_norm": 0.5565824508666992, + "learning_rate": 4.818601762772588e-05, + "loss": 0.4612, + "num_input_tokens_seen": 2149480, + "step": 8005 + }, + { + "epoch": 2.0996068152031455, + "grad_norm": 0.6884527206420898, + "learning_rate": 4.818173799836686e-05, + "loss": 0.6092, + "num_input_tokens_seen": 2150904, + "step": 8010 + }, + { + "epoch": 2.1009174311926606, + "grad_norm": 0.5675804615020752, + "learning_rate": 4.817745351711013e-05, + "loss": 0.3421, + "num_input_tokens_seen": 2152056, + "step": 8015 + }, + { + "epoch": 2.1022280471821757, + "grad_norm": 3.3155694007873535, + "learning_rate": 4.817316418485243e-05, + "loss": 0.5246, + "num_input_tokens_seen": 2153080, + "step": 8020 + }, + { + "epoch": 2.103538663171691, + "grad_norm": 0.4367039203643799, + "learning_rate": 4.81688700024915e-05, + "loss": 0.2316, + "num_input_tokens_seen": 2154312, + "step": 8025 + }, + { + "epoch": 2.104849279161206, + "grad_norm": 0.592885434627533, + "learning_rate": 4.816457097092611e-05, + "loss": 0.4507, + "num_input_tokens_seen": 2155528, + "step": 8030 + }, + { + "epoch": 2.106159895150721, + "grad_norm": 2.740217924118042, + "learning_rate": 4.816026709105605e-05, + "loss": 0.3278, + "num_input_tokens_seen": 2156568, + "step": 8035 + }, + { + "epoch": 2.107470511140236, + "grad_norm": 1.306555151939392, + "learning_rate": 4.81559583637821e-05, + "loss": 0.5457, + "num_input_tokens_seen": 2158056, + "step": 8040 + }, + { + "epoch": 2.108781127129751, + "grad_norm": 0.7905256748199463, + "learning_rate": 4.8151644790006077e-05, + "loss": 0.3783, + "num_input_tokens_seen": 2159160, + "step": 8045 + }, + { + "epoch": 2.1100917431192663, + "grad_norm": 0.8579105138778687, + "learning_rate": 4.8147326370630796e-05, + "loss": 0.4502, + "num_input_tokens_seen": 2160632, + "step": 8050 + }, + { + "epoch": 2.111402359108781, + "grad_norm": 1.5661057233810425, + "learning_rate": 4.814300310656011e-05, + "loss": 0.451, + "num_input_tokens_seen": 2162152, + "step": 8055 + }, + { + "epoch": 2.112712975098296, + "grad_norm": 0.8792870044708252, + "learning_rate": 4.8138674998698854e-05, + "loss": 0.5495, + "num_input_tokens_seen": 2163896, + "step": 8060 + }, + { + "epoch": 2.114023591087811, + "grad_norm": 0.760741114616394, + "learning_rate": 4.8134342047952906e-05, + "loss": 0.3351, + "num_input_tokens_seen": 2165240, + "step": 8065 + }, + { + "epoch": 2.1153342070773262, + "grad_norm": 0.35790562629699707, + "learning_rate": 4.8130004255229146e-05, + "loss": 0.3075, + "num_input_tokens_seen": 2166600, + "step": 8070 + }, + { + "epoch": 2.1166448230668413, + "grad_norm": 0.6542356014251709, + "learning_rate": 4.8125661621435445e-05, + "loss": 0.4641, + "num_input_tokens_seen": 2168056, + "step": 8075 + }, + { + "epoch": 2.1179554390563564, + "grad_norm": 0.9560678005218506, + "learning_rate": 4.8121314147480736e-05, + "loss": 0.356, + "num_input_tokens_seen": 2169480, + "step": 8080 + }, + { + "epoch": 2.1192660550458715, + "grad_norm": 0.5364248156547546, + "learning_rate": 4.811696183427493e-05, + "loss": 0.5278, + "num_input_tokens_seen": 2170632, + "step": 8085 + }, + { + "epoch": 2.1205766710353866, + "grad_norm": 0.8290550112724304, + "learning_rate": 4.811260468272895e-05, + "loss": 0.4615, + "num_input_tokens_seen": 2172072, + "step": 8090 + }, + { + "epoch": 2.1218872870249017, + "grad_norm": 0.6904206871986389, + "learning_rate": 4.810824269375475e-05, + "loss": 0.7033, + "num_input_tokens_seen": 2173416, + "step": 8095 + }, + { + "epoch": 2.123197903014417, + "grad_norm": 2.7157506942749023, + "learning_rate": 4.810387586826527e-05, + "loss": 0.5072, + "num_input_tokens_seen": 2174536, + "step": 8100 + }, + { + "epoch": 2.124508519003932, + "grad_norm": 0.6563311219215393, + "learning_rate": 4.80995042071745e-05, + "loss": 0.556, + "num_input_tokens_seen": 2176280, + "step": 8105 + }, + { + "epoch": 2.125819134993447, + "grad_norm": 0.6582857966423035, + "learning_rate": 4.809512771139741e-05, + "loss": 0.3431, + "num_input_tokens_seen": 2177928, + "step": 8110 + }, + { + "epoch": 2.127129750982962, + "grad_norm": 1.0994350910186768, + "learning_rate": 4.809074638184999e-05, + "loss": 0.3465, + "num_input_tokens_seen": 2179208, + "step": 8115 + }, + { + "epoch": 2.128440366972477, + "grad_norm": 1.0415223836898804, + "learning_rate": 4.808636021944927e-05, + "loss": 0.3543, + "num_input_tokens_seen": 2180520, + "step": 8120 + }, + { + "epoch": 2.1297509829619923, + "grad_norm": 1.1934638023376465, + "learning_rate": 4.808196922511323e-05, + "loss": 0.5503, + "num_input_tokens_seen": 2181800, + "step": 8125 + }, + { + "epoch": 2.1310615989515074, + "grad_norm": 1.817272424697876, + "learning_rate": 4.8077573399760914e-05, + "loss": 0.449, + "num_input_tokens_seen": 2183016, + "step": 8130 + }, + { + "epoch": 2.1323722149410225, + "grad_norm": 1.1467008590698242, + "learning_rate": 4.807317274431236e-05, + "loss": 0.4442, + "num_input_tokens_seen": 2184504, + "step": 8135 + }, + { + "epoch": 2.133682830930537, + "grad_norm": 1.3872236013412476, + "learning_rate": 4.806876725968862e-05, + "loss": 0.3541, + "num_input_tokens_seen": 2185544, + "step": 8140 + }, + { + "epoch": 2.134993446920052, + "grad_norm": 0.28148117661476135, + "learning_rate": 4.806435694681175e-05, + "loss": 0.2752, + "num_input_tokens_seen": 2187192, + "step": 8145 + }, + { + "epoch": 2.1363040629095673, + "grad_norm": 1.1991081237792969, + "learning_rate": 4.805994180660483e-05, + "loss": 0.4924, + "num_input_tokens_seen": 2188504, + "step": 8150 + }, + { + "epoch": 2.1376146788990824, + "grad_norm": 1.541317343711853, + "learning_rate": 4.805552183999193e-05, + "loss": 0.4521, + "num_input_tokens_seen": 2189928, + "step": 8155 + }, + { + "epoch": 2.1389252948885975, + "grad_norm": 1.5584977865219116, + "learning_rate": 4.8051097047898144e-05, + "loss": 0.5876, + "num_input_tokens_seen": 2191768, + "step": 8160 + }, + { + "epoch": 2.1402359108781126, + "grad_norm": 0.9716325402259827, + "learning_rate": 4.8046667431249585e-05, + "loss": 0.3982, + "num_input_tokens_seen": 2193080, + "step": 8165 + }, + { + "epoch": 2.1415465268676277, + "grad_norm": 0.49055346846580505, + "learning_rate": 4.804223299097334e-05, + "loss": 0.6818, + "num_input_tokens_seen": 2194104, + "step": 8170 + }, + { + "epoch": 2.142857142857143, + "grad_norm": 1.1350551843643188, + "learning_rate": 4.803779372799755e-05, + "loss": 0.4585, + "num_input_tokens_seen": 2197176, + "step": 8175 + }, + { + "epoch": 2.144167758846658, + "grad_norm": 1.7080296277999878, + "learning_rate": 4.803334964325134e-05, + "loss": 0.388, + "num_input_tokens_seen": 2198232, + "step": 8180 + }, + { + "epoch": 2.145478374836173, + "grad_norm": 0.537204921245575, + "learning_rate": 4.802890073766485e-05, + "loss": 0.4804, + "num_input_tokens_seen": 2199576, + "step": 8185 + }, + { + "epoch": 2.146788990825688, + "grad_norm": 0.978100597858429, + "learning_rate": 4.802444701216923e-05, + "loss": 0.7601, + "num_input_tokens_seen": 2200728, + "step": 8190 + }, + { + "epoch": 2.148099606815203, + "grad_norm": 0.5860198140144348, + "learning_rate": 4.801998846769662e-05, + "loss": 0.4281, + "num_input_tokens_seen": 2202296, + "step": 8195 + }, + { + "epoch": 2.1494102228047183, + "grad_norm": 0.9674468636512756, + "learning_rate": 4.801552510518021e-05, + "loss": 0.4946, + "num_input_tokens_seen": 2203480, + "step": 8200 + }, + { + "epoch": 2.1507208387942334, + "grad_norm": 0.679987370967865, + "learning_rate": 4.801105692555415e-05, + "loss": 0.4024, + "num_input_tokens_seen": 2204584, + "step": 8205 + }, + { + "epoch": 2.1520314547837485, + "grad_norm": 2.8428139686584473, + "learning_rate": 4.800658392975366e-05, + "loss": 0.4478, + "num_input_tokens_seen": 2205464, + "step": 8210 + }, + { + "epoch": 2.1533420707732636, + "grad_norm": 0.9571150541305542, + "learning_rate": 4.800210611871488e-05, + "loss": 0.6832, + "num_input_tokens_seen": 2206936, + "step": 8215 + }, + { + "epoch": 2.1546526867627787, + "grad_norm": 0.636827290058136, + "learning_rate": 4.7997623493375054e-05, + "loss": 0.4925, + "num_input_tokens_seen": 2208120, + "step": 8220 + }, + { + "epoch": 2.1559633027522938, + "grad_norm": 0.6871559023857117, + "learning_rate": 4.7993136054672356e-05, + "loss": 0.4825, + "num_input_tokens_seen": 2209608, + "step": 8225 + }, + { + "epoch": 2.157273918741809, + "grad_norm": 0.7749103903770447, + "learning_rate": 4.798864380354601e-05, + "loss": 0.2816, + "num_input_tokens_seen": 2211080, + "step": 8230 + }, + { + "epoch": 2.1585845347313235, + "grad_norm": 0.2953800857067108, + "learning_rate": 4.798414674093624e-05, + "loss": 0.5742, + "num_input_tokens_seen": 2212984, + "step": 8235 + }, + { + "epoch": 2.1598951507208386, + "grad_norm": 1.3387548923492432, + "learning_rate": 4.7979644867784276e-05, + "loss": 0.4363, + "num_input_tokens_seen": 2214136, + "step": 8240 + }, + { + "epoch": 2.1612057667103537, + "grad_norm": 4.814028739929199, + "learning_rate": 4.797513818503234e-05, + "loss": 0.5167, + "num_input_tokens_seen": 2215080, + "step": 8245 + }, + { + "epoch": 2.162516382699869, + "grad_norm": 1.3577678203582764, + "learning_rate": 4.797062669362369e-05, + "loss": 0.32, + "num_input_tokens_seen": 2216328, + "step": 8250 + }, + { + "epoch": 2.163826998689384, + "grad_norm": 0.8420731425285339, + "learning_rate": 4.796611039450255e-05, + "loss": 0.5996, + "num_input_tokens_seen": 2217608, + "step": 8255 + }, + { + "epoch": 2.165137614678899, + "grad_norm": 0.9473954439163208, + "learning_rate": 4.796158928861419e-05, + "loss": 0.4528, + "num_input_tokens_seen": 2218712, + "step": 8260 + }, + { + "epoch": 2.166448230668414, + "grad_norm": 1.2473444938659668, + "learning_rate": 4.7957063376904875e-05, + "loss": 0.3988, + "num_input_tokens_seen": 2219880, + "step": 8265 + }, + { + "epoch": 2.167758846657929, + "grad_norm": 0.9551470875740051, + "learning_rate": 4.7952532660321854e-05, + "loss": 0.5308, + "num_input_tokens_seen": 2221064, + "step": 8270 + }, + { + "epoch": 2.1690694626474443, + "grad_norm": 2.815207004547119, + "learning_rate": 4.794799713981341e-05, + "loss": 0.4302, + "num_input_tokens_seen": 2222440, + "step": 8275 + }, + { + "epoch": 2.1703800786369594, + "grad_norm": 0.5329520106315613, + "learning_rate": 4.7943456816328815e-05, + "loss": 0.4094, + "num_input_tokens_seen": 2223768, + "step": 8280 + }, + { + "epoch": 2.1716906946264745, + "grad_norm": 0.7793434858322144, + "learning_rate": 4.7938911690818347e-05, + "loss": 0.4225, + "num_input_tokens_seen": 2225320, + "step": 8285 + }, + { + "epoch": 2.1730013106159896, + "grad_norm": 1.370679259300232, + "learning_rate": 4.79343617642333e-05, + "loss": 0.3917, + "num_input_tokens_seen": 2226792, + "step": 8290 + }, + { + "epoch": 2.1743119266055047, + "grad_norm": 1.1209155321121216, + "learning_rate": 4.792980703752597e-05, + "loss": 0.4384, + "num_input_tokens_seen": 2228248, + "step": 8295 + }, + { + "epoch": 2.1756225425950197, + "grad_norm": 0.7044047713279724, + "learning_rate": 4.792524751164964e-05, + "loss": 0.3454, + "num_input_tokens_seen": 2229640, + "step": 8300 + }, + { + "epoch": 2.176933158584535, + "grad_norm": 0.7730234861373901, + "learning_rate": 4.792068318755861e-05, + "loss": 0.3094, + "num_input_tokens_seen": 2230936, + "step": 8305 + }, + { + "epoch": 2.17824377457405, + "grad_norm": 0.6042208075523376, + "learning_rate": 4.791611406620821e-05, + "loss": 0.4469, + "num_input_tokens_seen": 2232264, + "step": 8310 + }, + { + "epoch": 2.179554390563565, + "grad_norm": 0.7322703003883362, + "learning_rate": 4.7911540148554725e-05, + "loss": 0.3739, + "num_input_tokens_seen": 2233416, + "step": 8315 + }, + { + "epoch": 2.18086500655308, + "grad_norm": 0.7189313173294067, + "learning_rate": 4.790696143555546e-05, + "loss": 0.4151, + "num_input_tokens_seen": 2235032, + "step": 8320 + }, + { + "epoch": 2.1821756225425952, + "grad_norm": 0.9266195297241211, + "learning_rate": 4.790237792816877e-05, + "loss": 0.8006, + "num_input_tokens_seen": 2236440, + "step": 8325 + }, + { + "epoch": 2.18348623853211, + "grad_norm": 0.4843691885471344, + "learning_rate": 4.789778962735394e-05, + "loss": 0.3095, + "num_input_tokens_seen": 2237912, + "step": 8330 + }, + { + "epoch": 2.184796854521625, + "grad_norm": 2.427912712097168, + "learning_rate": 4.78931965340713e-05, + "loss": 0.4398, + "num_input_tokens_seen": 2239144, + "step": 8335 + }, + { + "epoch": 2.18610747051114, + "grad_norm": 1.3838759660720825, + "learning_rate": 4.78885986492822e-05, + "loss": 0.5023, + "num_input_tokens_seen": 2240840, + "step": 8340 + }, + { + "epoch": 2.187418086500655, + "grad_norm": 0.842789888381958, + "learning_rate": 4.788399597394894e-05, + "loss": 0.4935, + "num_input_tokens_seen": 2242296, + "step": 8345 + }, + { + "epoch": 2.1887287024901703, + "grad_norm": 0.9258467555046082, + "learning_rate": 4.787938850903486e-05, + "loss": 0.4482, + "num_input_tokens_seen": 2243608, + "step": 8350 + }, + { + "epoch": 2.1900393184796854, + "grad_norm": 1.3631362915039062, + "learning_rate": 4.78747762555043e-05, + "loss": 0.5492, + "num_input_tokens_seen": 2244888, + "step": 8355 + }, + { + "epoch": 2.1913499344692005, + "grad_norm": 0.7516872882843018, + "learning_rate": 4.7870159214322594e-05, + "loss": 0.421, + "num_input_tokens_seen": 2246136, + "step": 8360 + }, + { + "epoch": 2.1926605504587156, + "grad_norm": 1.5895628929138184, + "learning_rate": 4.7865537386456084e-05, + "loss": 0.4918, + "num_input_tokens_seen": 2247480, + "step": 8365 + }, + { + "epoch": 2.1939711664482306, + "grad_norm": 0.5233879089355469, + "learning_rate": 4.7860910772872104e-05, + "loss": 0.337, + "num_input_tokens_seen": 2248456, + "step": 8370 + }, + { + "epoch": 2.1952817824377457, + "grad_norm": 1.0205384492874146, + "learning_rate": 4.7856279374539e-05, + "loss": 0.752, + "num_input_tokens_seen": 2249608, + "step": 8375 + }, + { + "epoch": 2.196592398427261, + "grad_norm": 2.3728902339935303, + "learning_rate": 4.785164319242612e-05, + "loss": 0.4285, + "num_input_tokens_seen": 2250648, + "step": 8380 + }, + { + "epoch": 2.197903014416776, + "grad_norm": 0.5376797318458557, + "learning_rate": 4.78470022275038e-05, + "loss": 0.6134, + "num_input_tokens_seen": 2252168, + "step": 8385 + }, + { + "epoch": 2.199213630406291, + "grad_norm": 0.5806631445884705, + "learning_rate": 4.784235648074339e-05, + "loss": 0.2692, + "num_input_tokens_seen": 2253464, + "step": 8390 + }, + { + "epoch": 2.200524246395806, + "grad_norm": 1.2106074094772339, + "learning_rate": 4.783770595311724e-05, + "loss": 0.3533, + "num_input_tokens_seen": 2254712, + "step": 8395 + }, + { + "epoch": 2.2018348623853212, + "grad_norm": 0.38610053062438965, + "learning_rate": 4.78330506455987e-05, + "loss": 0.2886, + "num_input_tokens_seen": 2255800, + "step": 8400 + }, + { + "epoch": 2.2031454783748363, + "grad_norm": 0.3355385959148407, + "learning_rate": 4.78283905591621e-05, + "loss": 0.4786, + "num_input_tokens_seen": 2257208, + "step": 8405 + }, + { + "epoch": 2.2044560943643514, + "grad_norm": 0.9646286964416504, + "learning_rate": 4.78237256947828e-05, + "loss": 0.441, + "num_input_tokens_seen": 2258600, + "step": 8410 + }, + { + "epoch": 2.2057667103538665, + "grad_norm": 0.5593180656433105, + "learning_rate": 4.7819056053437165e-05, + "loss": 0.5211, + "num_input_tokens_seen": 2259960, + "step": 8415 + }, + { + "epoch": 2.2070773263433816, + "grad_norm": 0.8802770376205444, + "learning_rate": 4.7814381636102515e-05, + "loss": 0.5039, + "num_input_tokens_seen": 2260888, + "step": 8420 + }, + { + "epoch": 2.2083879423328963, + "grad_norm": 0.6311805248260498, + "learning_rate": 4.780970244375721e-05, + "loss": 0.4908, + "num_input_tokens_seen": 2262376, + "step": 8425 + }, + { + "epoch": 2.2096985583224114, + "grad_norm": 0.47813668847084045, + "learning_rate": 4.780501847738059e-05, + "loss": 0.4299, + "num_input_tokens_seen": 2263432, + "step": 8430 + }, + { + "epoch": 2.2110091743119265, + "grad_norm": 0.6305199265480042, + "learning_rate": 4.7800329737953016e-05, + "loss": 0.4216, + "num_input_tokens_seen": 2264808, + "step": 8435 + }, + { + "epoch": 2.2123197903014415, + "grad_norm": 0.7314751148223877, + "learning_rate": 4.779563622645582e-05, + "loss": 0.4388, + "num_input_tokens_seen": 2266248, + "step": 8440 + }, + { + "epoch": 2.2136304062909566, + "grad_norm": 0.9380919933319092, + "learning_rate": 4.779093794387135e-05, + "loss": 0.4034, + "num_input_tokens_seen": 2267416, + "step": 8445 + }, + { + "epoch": 2.2149410222804717, + "grad_norm": 0.6142908930778503, + "learning_rate": 4.778623489118295e-05, + "loss": 0.4313, + "num_input_tokens_seen": 2268440, + "step": 8450 + }, + { + "epoch": 2.216251638269987, + "grad_norm": 2.332784652709961, + "learning_rate": 4.7781527069374957e-05, + "loss": 0.4603, + "num_input_tokens_seen": 2269800, + "step": 8455 + }, + { + "epoch": 2.217562254259502, + "grad_norm": 0.6008874773979187, + "learning_rate": 4.777681447943271e-05, + "loss": 0.4566, + "num_input_tokens_seen": 2271256, + "step": 8460 + }, + { + "epoch": 2.218872870249017, + "grad_norm": 0.7660607695579529, + "learning_rate": 4.777209712234255e-05, + "loss": 0.3461, + "num_input_tokens_seen": 2272728, + "step": 8465 + }, + { + "epoch": 2.220183486238532, + "grad_norm": 0.6393540501594543, + "learning_rate": 4.776737499909181e-05, + "loss": 0.7434, + "num_input_tokens_seen": 2274072, + "step": 8470 + }, + { + "epoch": 2.221494102228047, + "grad_norm": 0.48751428723335266, + "learning_rate": 4.776264811066882e-05, + "loss": 0.4818, + "num_input_tokens_seen": 2275560, + "step": 8475 + }, + { + "epoch": 2.2228047182175623, + "grad_norm": 0.3206268846988678, + "learning_rate": 4.7757916458062905e-05, + "loss": 0.3445, + "num_input_tokens_seen": 2276840, + "step": 8480 + }, + { + "epoch": 2.2241153342070774, + "grad_norm": 0.8750942945480347, + "learning_rate": 4.7753180042264404e-05, + "loss": 0.6079, + "num_input_tokens_seen": 2277944, + "step": 8485 + }, + { + "epoch": 2.2254259501965925, + "grad_norm": 2.0172524452209473, + "learning_rate": 4.7748438864264636e-05, + "loss": 0.3587, + "num_input_tokens_seen": 2278792, + "step": 8490 + }, + { + "epoch": 2.2267365661861076, + "grad_norm": 1.368959903717041, + "learning_rate": 4.774369292505592e-05, + "loss": 0.3962, + "num_input_tokens_seen": 2280488, + "step": 8495 + }, + { + "epoch": 2.2280471821756227, + "grad_norm": 0.5842744708061218, + "learning_rate": 4.773894222563157e-05, + "loss": 0.4009, + "num_input_tokens_seen": 2281640, + "step": 8500 + }, + { + "epoch": 2.229357798165138, + "grad_norm": 0.8206716775894165, + "learning_rate": 4.7734186766985904e-05, + "loss": 0.3692, + "num_input_tokens_seen": 2282744, + "step": 8505 + }, + { + "epoch": 2.230668414154653, + "grad_norm": 0.6643772125244141, + "learning_rate": 4.772942655011423e-05, + "loss": 0.2417, + "num_input_tokens_seen": 2283848, + "step": 8510 + }, + { + "epoch": 2.231979030144168, + "grad_norm": 1.3837509155273438, + "learning_rate": 4.772466157601285e-05, + "loss": 0.4677, + "num_input_tokens_seen": 2285000, + "step": 8515 + }, + { + "epoch": 2.2332896461336826, + "grad_norm": 0.8118686676025391, + "learning_rate": 4.771989184567907e-05, + "loss": 0.3451, + "num_input_tokens_seen": 2286056, + "step": 8520 + }, + { + "epoch": 2.2346002621231977, + "grad_norm": 1.2496055364608765, + "learning_rate": 4.7715117360111185e-05, + "loss": 0.6986, + "num_input_tokens_seen": 2287320, + "step": 8525 + }, + { + "epoch": 2.235910878112713, + "grad_norm": 1.069921612739563, + "learning_rate": 4.771033812030849e-05, + "loss": 0.4901, + "num_input_tokens_seen": 2288280, + "step": 8530 + }, + { + "epoch": 2.237221494102228, + "grad_norm": 2.0298256874084473, + "learning_rate": 4.7705554127271254e-05, + "loss": 0.4383, + "num_input_tokens_seen": 2289576, + "step": 8535 + }, + { + "epoch": 2.238532110091743, + "grad_norm": 0.698921799659729, + "learning_rate": 4.770076538200078e-05, + "loss": 0.3268, + "num_input_tokens_seen": 2291032, + "step": 8540 + }, + { + "epoch": 2.239842726081258, + "grad_norm": 0.5877149701118469, + "learning_rate": 4.769597188549932e-05, + "loss": 0.4504, + "num_input_tokens_seen": 2292824, + "step": 8545 + }, + { + "epoch": 2.241153342070773, + "grad_norm": 0.5182098150253296, + "learning_rate": 4.769117363877018e-05, + "loss": 0.4152, + "num_input_tokens_seen": 2294264, + "step": 8550 + }, + { + "epoch": 2.2424639580602883, + "grad_norm": 0.6673553586006165, + "learning_rate": 4.76863706428176e-05, + "loss": 0.5358, + "num_input_tokens_seen": 2295960, + "step": 8555 + }, + { + "epoch": 2.2437745740498034, + "grad_norm": 1.0411354303359985, + "learning_rate": 4.7681562898646836e-05, + "loss": 0.4933, + "num_input_tokens_seen": 2297336, + "step": 8560 + }, + { + "epoch": 2.2450851900393185, + "grad_norm": 0.9386358857154846, + "learning_rate": 4.767675040726416e-05, + "loss": 0.3728, + "num_input_tokens_seen": 2298568, + "step": 8565 + }, + { + "epoch": 2.2463958060288336, + "grad_norm": 0.33773675560951233, + "learning_rate": 4.76719331696768e-05, + "loss": 0.4515, + "num_input_tokens_seen": 2299576, + "step": 8570 + }, + { + "epoch": 2.2477064220183487, + "grad_norm": 0.8440136909484863, + "learning_rate": 4.7667111186893e-05, + "loss": 0.5322, + "num_input_tokens_seen": 2301288, + "step": 8575 + }, + { + "epoch": 2.249017038007864, + "grad_norm": 0.3354502320289612, + "learning_rate": 4.766228445992199e-05, + "loss": 0.3726, + "num_input_tokens_seen": 2302664, + "step": 8580 + }, + { + "epoch": 2.250327653997379, + "grad_norm": 0.6033670902252197, + "learning_rate": 4.7657452989774e-05, + "loss": 0.3679, + "num_input_tokens_seen": 2303960, + "step": 8585 + }, + { + "epoch": 2.251638269986894, + "grad_norm": 0.6323764324188232, + "learning_rate": 4.765261677746026e-05, + "loss": 0.3822, + "num_input_tokens_seen": 2305112, + "step": 8590 + }, + { + "epoch": 2.252948885976409, + "grad_norm": 0.5209841132164001, + "learning_rate": 4.764777582399296e-05, + "loss": 0.3362, + "num_input_tokens_seen": 2306600, + "step": 8595 + }, + { + "epoch": 2.254259501965924, + "grad_norm": 1.6461353302001953, + "learning_rate": 4.764293013038531e-05, + "loss": 0.4482, + "num_input_tokens_seen": 2307928, + "step": 8600 + }, + { + "epoch": 2.255570117955439, + "grad_norm": 0.5541807413101196, + "learning_rate": 4.7638079697651514e-05, + "loss": 0.5352, + "num_input_tokens_seen": 2309144, + "step": 8605 + }, + { + "epoch": 2.2568807339449544, + "grad_norm": 0.7290766835212708, + "learning_rate": 4.763322452680675e-05, + "loss": 0.379, + "num_input_tokens_seen": 2310296, + "step": 8610 + }, + { + "epoch": 2.258191349934469, + "grad_norm": 0.42221692204475403, + "learning_rate": 4.76283646188672e-05, + "loss": 0.3403, + "num_input_tokens_seen": 2311816, + "step": 8615 + }, + { + "epoch": 2.259501965923984, + "grad_norm": 0.6365774273872375, + "learning_rate": 4.7623499974850036e-05, + "loss": 0.4279, + "num_input_tokens_seen": 2313112, + "step": 8620 + }, + { + "epoch": 2.260812581913499, + "grad_norm": 0.7125449180603027, + "learning_rate": 4.7618630595773424e-05, + "loss": 0.4525, + "num_input_tokens_seen": 2314312, + "step": 8625 + }, + { + "epoch": 2.2621231979030143, + "grad_norm": 1.9527770280838013, + "learning_rate": 4.7613756482656505e-05, + "loss": 0.6885, + "num_input_tokens_seen": 2315432, + "step": 8630 + }, + { + "epoch": 2.2634338138925294, + "grad_norm": 0.4094899594783783, + "learning_rate": 4.760887763651942e-05, + "loss": 0.3807, + "num_input_tokens_seen": 2317128, + "step": 8635 + }, + { + "epoch": 2.2647444298820445, + "grad_norm": 0.7568178176879883, + "learning_rate": 4.7603994058383315e-05, + "loss": 0.4486, + "num_input_tokens_seen": 2318904, + "step": 8640 + }, + { + "epoch": 2.2660550458715596, + "grad_norm": 0.6729658246040344, + "learning_rate": 4.759910574927032e-05, + "loss": 0.5333, + "num_input_tokens_seen": 2321384, + "step": 8645 + }, + { + "epoch": 2.2673656618610747, + "grad_norm": 0.7405988574028015, + "learning_rate": 4.759421271020353e-05, + "loss": 0.3772, + "num_input_tokens_seen": 2322776, + "step": 8650 + }, + { + "epoch": 2.26867627785059, + "grad_norm": 0.6106330156326294, + "learning_rate": 4.758931494220707e-05, + "loss": 0.3439, + "num_input_tokens_seen": 2324376, + "step": 8655 + }, + { + "epoch": 2.269986893840105, + "grad_norm": 0.5513662099838257, + "learning_rate": 4.7584412446306e-05, + "loss": 0.4738, + "num_input_tokens_seen": 2325688, + "step": 8660 + }, + { + "epoch": 2.27129750982962, + "grad_norm": 0.629091203212738, + "learning_rate": 4.7579505223526446e-05, + "loss": 0.5997, + "num_input_tokens_seen": 2326808, + "step": 8665 + }, + { + "epoch": 2.272608125819135, + "grad_norm": 0.4248215854167938, + "learning_rate": 4.757459327489546e-05, + "loss": 0.2558, + "num_input_tokens_seen": 2328344, + "step": 8670 + }, + { + "epoch": 2.27391874180865, + "grad_norm": 0.948695719242096, + "learning_rate": 4.756967660144111e-05, + "loss": 0.3132, + "num_input_tokens_seen": 2329560, + "step": 8675 + }, + { + "epoch": 2.2752293577981653, + "grad_norm": 0.47379133105278015, + "learning_rate": 4.756475520419244e-05, + "loss": 0.5607, + "num_input_tokens_seen": 2330888, + "step": 8680 + }, + { + "epoch": 2.2765399737876804, + "grad_norm": 1.4888899326324463, + "learning_rate": 4.7559829084179494e-05, + "loss": 0.6483, + "num_input_tokens_seen": 2332232, + "step": 8685 + }, + { + "epoch": 2.2778505897771955, + "grad_norm": 0.5310717821121216, + "learning_rate": 4.755489824243331e-05, + "loss": 0.7829, + "num_input_tokens_seen": 2333512, + "step": 8690 + }, + { + "epoch": 2.2791612057667106, + "grad_norm": 0.36532196402549744, + "learning_rate": 4.754996267998589e-05, + "loss": 0.4363, + "num_input_tokens_seen": 2334792, + "step": 8695 + }, + { + "epoch": 2.280471821756225, + "grad_norm": 0.7704387903213501, + "learning_rate": 4.754502239787024e-05, + "loss": 0.3812, + "num_input_tokens_seen": 2336040, + "step": 8700 + }, + { + "epoch": 2.2817824377457407, + "grad_norm": 1.8028831481933594, + "learning_rate": 4.7540077397120365e-05, + "loss": 0.5371, + "num_input_tokens_seen": 2336936, + "step": 8705 + }, + { + "epoch": 2.2830930537352554, + "grad_norm": 0.6338413953781128, + "learning_rate": 4.7535127678771233e-05, + "loss": 0.5389, + "num_input_tokens_seen": 2338664, + "step": 8710 + }, + { + "epoch": 2.2844036697247705, + "grad_norm": 0.7639115452766418, + "learning_rate": 4.7530173243858814e-05, + "loss": 0.4584, + "num_input_tokens_seen": 2339816, + "step": 8715 + }, + { + "epoch": 2.2857142857142856, + "grad_norm": 0.9075050950050354, + "learning_rate": 4.7525214093420065e-05, + "loss": 0.4336, + "num_input_tokens_seen": 2341128, + "step": 8720 + }, + { + "epoch": 2.2870249017038007, + "grad_norm": 0.6908093690872192, + "learning_rate": 4.7520250228492934e-05, + "loss": 0.3475, + "num_input_tokens_seen": 2342184, + "step": 8725 + }, + { + "epoch": 2.288335517693316, + "grad_norm": 0.7397432923316956, + "learning_rate": 4.751528165011633e-05, + "loss": 0.3505, + "num_input_tokens_seen": 2343960, + "step": 8730 + }, + { + "epoch": 2.289646133682831, + "grad_norm": 1.4229596853256226, + "learning_rate": 4.7510308359330194e-05, + "loss": 0.3953, + "num_input_tokens_seen": 2345208, + "step": 8735 + }, + { + "epoch": 2.290956749672346, + "grad_norm": 0.8584851622581482, + "learning_rate": 4.7505330357175406e-05, + "loss": 0.5168, + "num_input_tokens_seen": 2346568, + "step": 8740 + }, + { + "epoch": 2.292267365661861, + "grad_norm": 1.2250618934631348, + "learning_rate": 4.750034764469386e-05, + "loss": 0.4923, + "num_input_tokens_seen": 2347656, + "step": 8745 + }, + { + "epoch": 2.293577981651376, + "grad_norm": 0.45459508895874023, + "learning_rate": 4.749536022292842e-05, + "loss": 0.4047, + "num_input_tokens_seen": 2349112, + "step": 8750 + }, + { + "epoch": 2.2948885976408913, + "grad_norm": 0.5099465250968933, + "learning_rate": 4.7490368092922964e-05, + "loss": 0.3139, + "num_input_tokens_seen": 2350536, + "step": 8755 + }, + { + "epoch": 2.2961992136304064, + "grad_norm": 0.4704533517360687, + "learning_rate": 4.748537125572233e-05, + "loss": 0.7901, + "num_input_tokens_seen": 2352472, + "step": 8760 + }, + { + "epoch": 2.2975098296199215, + "grad_norm": 4.823760032653809, + "learning_rate": 4.748036971237234e-05, + "loss": 0.6527, + "num_input_tokens_seen": 2353784, + "step": 8765 + }, + { + "epoch": 2.2988204456094365, + "grad_norm": 1.7508286237716675, + "learning_rate": 4.747536346391981e-05, + "loss": 0.3771, + "num_input_tokens_seen": 2355240, + "step": 8770 + }, + { + "epoch": 2.3001310615989516, + "grad_norm": 0.5960665941238403, + "learning_rate": 4.747035251141254e-05, + "loss": 0.3774, + "num_input_tokens_seen": 2356312, + "step": 8775 + }, + { + "epoch": 2.3014416775884667, + "grad_norm": 2.8373665809631348, + "learning_rate": 4.746533685589932e-05, + "loss": 0.3588, + "num_input_tokens_seen": 2357624, + "step": 8780 + }, + { + "epoch": 2.302752293577982, + "grad_norm": 10.495039939880371, + "learning_rate": 4.746031649842991e-05, + "loss": 0.4801, + "num_input_tokens_seen": 2358984, + "step": 8785 + }, + { + "epoch": 2.304062909567497, + "grad_norm": 0.7929167151451111, + "learning_rate": 4.745529144005507e-05, + "loss": 0.3515, + "num_input_tokens_seen": 2360056, + "step": 8790 + }, + { + "epoch": 2.3053735255570116, + "grad_norm": 0.8963240385055542, + "learning_rate": 4.745026168182654e-05, + "loss": 0.7031, + "num_input_tokens_seen": 2363800, + "step": 8795 + }, + { + "epoch": 2.3066841415465267, + "grad_norm": 0.4707227349281311, + "learning_rate": 4.744522722479701e-05, + "loss": 0.4003, + "num_input_tokens_seen": 2365224, + "step": 8800 + }, + { + "epoch": 2.3079947575360418, + "grad_norm": 0.6832383871078491, + "learning_rate": 4.7440188070020217e-05, + "loss": 0.617, + "num_input_tokens_seen": 2366840, + "step": 8805 + }, + { + "epoch": 2.309305373525557, + "grad_norm": 0.6102614998817444, + "learning_rate": 4.7435144218550834e-05, + "loss": 0.3769, + "num_input_tokens_seen": 2368424, + "step": 8810 + }, + { + "epoch": 2.310615989515072, + "grad_norm": 0.7898614406585693, + "learning_rate": 4.7430095671444526e-05, + "loss": 1.0062, + "num_input_tokens_seen": 2370024, + "step": 8815 + }, + { + "epoch": 2.311926605504587, + "grad_norm": 1.2592105865478516, + "learning_rate": 4.742504242975795e-05, + "loss": 0.3014, + "num_input_tokens_seen": 2371448, + "step": 8820 + }, + { + "epoch": 2.313237221494102, + "grad_norm": 0.8522607088088989, + "learning_rate": 4.7419984494548744e-05, + "loss": 0.5676, + "num_input_tokens_seen": 2373080, + "step": 8825 + }, + { + "epoch": 2.3145478374836173, + "grad_norm": 2.033315896987915, + "learning_rate": 4.7414921866875524e-05, + "loss": 0.5342, + "num_input_tokens_seen": 2374088, + "step": 8830 + }, + { + "epoch": 2.3158584534731324, + "grad_norm": 0.7666423916816711, + "learning_rate": 4.740985454779788e-05, + "loss": 0.4104, + "num_input_tokens_seen": 2375848, + "step": 8835 + }, + { + "epoch": 2.3171690694626474, + "grad_norm": 0.8793627023696899, + "learning_rate": 4.740478253837641e-05, + "loss": 0.3652, + "num_input_tokens_seen": 2376920, + "step": 8840 + }, + { + "epoch": 2.3184796854521625, + "grad_norm": 1.6911104917526245, + "learning_rate": 4.739970583967266e-05, + "loss": 0.5764, + "num_input_tokens_seen": 2378584, + "step": 8845 + }, + { + "epoch": 2.3197903014416776, + "grad_norm": 0.9970225691795349, + "learning_rate": 4.739462445274919e-05, + "loss": 0.6584, + "num_input_tokens_seen": 2379848, + "step": 8850 + }, + { + "epoch": 2.3211009174311927, + "grad_norm": 0.7615451216697693, + "learning_rate": 4.738953837866951e-05, + "loss": 0.418, + "num_input_tokens_seen": 2381048, + "step": 8855 + }, + { + "epoch": 2.322411533420708, + "grad_norm": 0.7397096157073975, + "learning_rate": 4.738444761849813e-05, + "loss": 0.3726, + "num_input_tokens_seen": 2382232, + "step": 8860 + }, + { + "epoch": 2.323722149410223, + "grad_norm": 0.9014339447021484, + "learning_rate": 4.7379352173300553e-05, + "loss": 0.5525, + "num_input_tokens_seen": 2383368, + "step": 8865 + }, + { + "epoch": 2.325032765399738, + "grad_norm": 0.9001389741897583, + "learning_rate": 4.737425204414322e-05, + "loss": 0.409, + "num_input_tokens_seen": 2385032, + "step": 8870 + }, + { + "epoch": 2.326343381389253, + "grad_norm": 0.653905987739563, + "learning_rate": 4.73691472320936e-05, + "loss": 0.2479, + "num_input_tokens_seen": 2388264, + "step": 8875 + }, + { + "epoch": 2.327653997378768, + "grad_norm": 1.3812716007232666, + "learning_rate": 4.736403773822012e-05, + "loss": 0.4851, + "num_input_tokens_seen": 2389464, + "step": 8880 + }, + { + "epoch": 2.3289646133682833, + "grad_norm": 1.4066905975341797, + "learning_rate": 4.735892356359217e-05, + "loss": 0.4147, + "num_input_tokens_seen": 2390456, + "step": 8885 + }, + { + "epoch": 2.330275229357798, + "grad_norm": 1.1680946350097656, + "learning_rate": 4.735380470928017e-05, + "loss": 0.5938, + "num_input_tokens_seen": 2392168, + "step": 8890 + }, + { + "epoch": 2.331585845347313, + "grad_norm": 0.6025600433349609, + "learning_rate": 4.7348681176355456e-05, + "loss": 0.3575, + "num_input_tokens_seen": 2393464, + "step": 8895 + }, + { + "epoch": 2.332896461336828, + "grad_norm": 0.48119187355041504, + "learning_rate": 4.734355296589039e-05, + "loss": 0.6183, + "num_input_tokens_seen": 2395064, + "step": 8900 + }, + { + "epoch": 2.3342070773263432, + "grad_norm": 0.4041799008846283, + "learning_rate": 4.73384200789583e-05, + "loss": 0.4551, + "num_input_tokens_seen": 2396232, + "step": 8905 + }, + { + "epoch": 2.3355176933158583, + "grad_norm": 1.419061303138733, + "learning_rate": 4.733328251663348e-05, + "loss": 0.4187, + "num_input_tokens_seen": 2397544, + "step": 8910 + }, + { + "epoch": 2.3368283093053734, + "grad_norm": 0.779507040977478, + "learning_rate": 4.732814027999122e-05, + "loss": 0.3438, + "num_input_tokens_seen": 2398824, + "step": 8915 + }, + { + "epoch": 2.3381389252948885, + "grad_norm": 0.8640726208686829, + "learning_rate": 4.7322993370107775e-05, + "loss": 0.6503, + "num_input_tokens_seen": 2399816, + "step": 8920 + }, + { + "epoch": 2.3394495412844036, + "grad_norm": 0.7747657299041748, + "learning_rate": 4.7317841788060394e-05, + "loss": 0.4588, + "num_input_tokens_seen": 2400952, + "step": 8925 + }, + { + "epoch": 2.3407601572739187, + "grad_norm": 0.9455310702323914, + "learning_rate": 4.731268553492729e-05, + "loss": 0.3626, + "num_input_tokens_seen": 2402120, + "step": 8930 + }, + { + "epoch": 2.342070773263434, + "grad_norm": 0.38150501251220703, + "learning_rate": 4.730752461178766e-05, + "loss": 0.3758, + "num_input_tokens_seen": 2403432, + "step": 8935 + }, + { + "epoch": 2.343381389252949, + "grad_norm": 0.43593040108680725, + "learning_rate": 4.730235901972166e-05, + "loss": 0.4225, + "num_input_tokens_seen": 2405160, + "step": 8940 + }, + { + "epoch": 2.344692005242464, + "grad_norm": 0.4706116020679474, + "learning_rate": 4.729718875981046e-05, + "loss": 0.3222, + "num_input_tokens_seen": 2406536, + "step": 8945 + }, + { + "epoch": 2.346002621231979, + "grad_norm": 1.0756194591522217, + "learning_rate": 4.729201383313618e-05, + "loss": 0.5067, + "num_input_tokens_seen": 2407992, + "step": 8950 + }, + { + "epoch": 2.347313237221494, + "grad_norm": 0.9922107458114624, + "learning_rate": 4.7286834240781925e-05, + "loss": 0.2141, + "num_input_tokens_seen": 2409016, + "step": 8955 + }, + { + "epoch": 2.3486238532110093, + "grad_norm": 1.6996794939041138, + "learning_rate": 4.728164998383177e-05, + "loss": 0.5025, + "num_input_tokens_seen": 2410472, + "step": 8960 + }, + { + "epoch": 2.3499344692005244, + "grad_norm": 0.7564626336097717, + "learning_rate": 4.727646106337077e-05, + "loss": 0.3503, + "num_input_tokens_seen": 2411496, + "step": 8965 + }, + { + "epoch": 2.3512450851900395, + "grad_norm": 1.1213126182556152, + "learning_rate": 4.727126748048496e-05, + "loss": 0.4047, + "num_input_tokens_seen": 2412696, + "step": 8970 + }, + { + "epoch": 2.352555701179554, + "grad_norm": 0.8002477884292603, + "learning_rate": 4.726606923626135e-05, + "loss": 0.3564, + "num_input_tokens_seen": 2414008, + "step": 8975 + }, + { + "epoch": 2.3538663171690697, + "grad_norm": 0.9117506146430969, + "learning_rate": 4.726086633178792e-05, + "loss": 0.3402, + "num_input_tokens_seen": 2415272, + "step": 8980 + }, + { + "epoch": 2.3551769331585843, + "grad_norm": 1.3982261419296265, + "learning_rate": 4.7255658768153635e-05, + "loss": 0.5096, + "num_input_tokens_seen": 2417240, + "step": 8985 + }, + { + "epoch": 2.3564875491480994, + "grad_norm": 1.8548985719680786, + "learning_rate": 4.725044654644842e-05, + "loss": 0.5479, + "num_input_tokens_seen": 2418376, + "step": 8990 + }, + { + "epoch": 2.3577981651376145, + "grad_norm": 0.5279331803321838, + "learning_rate": 4.724522966776319e-05, + "loss": 0.4134, + "num_input_tokens_seen": 2420120, + "step": 8995 + }, + { + "epoch": 2.3591087811271296, + "grad_norm": 1.3607385158538818, + "learning_rate": 4.7240008133189836e-05, + "loss": 0.4695, + "num_input_tokens_seen": 2421160, + "step": 9000 + }, + { + "epoch": 2.3604193971166447, + "grad_norm": 0.6538021564483643, + "learning_rate": 4.723478194382119e-05, + "loss": 0.3422, + "num_input_tokens_seen": 2423048, + "step": 9005 + }, + { + "epoch": 2.36173001310616, + "grad_norm": 1.165535569190979, + "learning_rate": 4.722955110075112e-05, + "loss": 0.3498, + "num_input_tokens_seen": 2424504, + "step": 9010 + }, + { + "epoch": 2.363040629095675, + "grad_norm": 1.6095051765441895, + "learning_rate": 4.7224315605074403e-05, + "loss": 0.3784, + "num_input_tokens_seen": 2425416, + "step": 9015 + }, + { + "epoch": 2.36435124508519, + "grad_norm": 0.8876811861991882, + "learning_rate": 4.7219075457886836e-05, + "loss": 0.3864, + "num_input_tokens_seen": 2427176, + "step": 9020 + }, + { + "epoch": 2.365661861074705, + "grad_norm": 0.8176876306533813, + "learning_rate": 4.721383066028516e-05, + "loss": 0.5242, + "num_input_tokens_seen": 2428360, + "step": 9025 + }, + { + "epoch": 2.36697247706422, + "grad_norm": 0.5918155908584595, + "learning_rate": 4.720858121336712e-05, + "loss": 0.3506, + "num_input_tokens_seen": 2429688, + "step": 9030 + }, + { + "epoch": 2.3682830930537353, + "grad_norm": 1.1445131301879883, + "learning_rate": 4.72033271182314e-05, + "loss": 0.4718, + "num_input_tokens_seen": 2430712, + "step": 9035 + }, + { + "epoch": 2.3695937090432504, + "grad_norm": 1.0759609937667847, + "learning_rate": 4.7198068375977675e-05, + "loss": 0.476, + "num_input_tokens_seen": 2432072, + "step": 9040 + }, + { + "epoch": 2.3709043250327655, + "grad_norm": 0.48665088415145874, + "learning_rate": 4.71928049877066e-05, + "loss": 0.3583, + "num_input_tokens_seen": 2433288, + "step": 9045 + }, + { + "epoch": 2.3722149410222806, + "grad_norm": 1.3783577680587769, + "learning_rate": 4.718753695451978e-05, + "loss": 0.3707, + "num_input_tokens_seen": 2434616, + "step": 9050 + }, + { + "epoch": 2.3735255570117957, + "grad_norm": 0.7773056626319885, + "learning_rate": 4.718226427751982e-05, + "loss": 0.5526, + "num_input_tokens_seen": 2435688, + "step": 9055 + }, + { + "epoch": 2.374836173001311, + "grad_norm": 0.5385114550590515, + "learning_rate": 4.7176986957810275e-05, + "loss": 0.4818, + "num_input_tokens_seen": 2437416, + "step": 9060 + }, + { + "epoch": 2.376146788990826, + "grad_norm": 0.839625895023346, + "learning_rate": 4.717170499649568e-05, + "loss": 0.4802, + "num_input_tokens_seen": 2438280, + "step": 9065 + }, + { + "epoch": 2.3774574049803405, + "grad_norm": 1.6452292203903198, + "learning_rate": 4.7166418394681526e-05, + "loss": 0.5252, + "num_input_tokens_seen": 2439784, + "step": 9070 + }, + { + "epoch": 2.378768020969856, + "grad_norm": 1.0466471910476685, + "learning_rate": 4.716112715347432e-05, + "loss": 0.3802, + "num_input_tokens_seen": 2441160, + "step": 9075 + }, + { + "epoch": 2.3800786369593707, + "grad_norm": 0.5291253328323364, + "learning_rate": 4.715583127398148e-05, + "loss": 0.4147, + "num_input_tokens_seen": 2442328, + "step": 9080 + }, + { + "epoch": 2.381389252948886, + "grad_norm": 1.0118540525436401, + "learning_rate": 4.715053075731144e-05, + "loss": 0.2599, + "num_input_tokens_seen": 2443432, + "step": 9085 + }, + { + "epoch": 2.382699868938401, + "grad_norm": 0.7866188287734985, + "learning_rate": 4.714522560457359e-05, + "loss": 0.4452, + "num_input_tokens_seen": 2444488, + "step": 9090 + }, + { + "epoch": 2.384010484927916, + "grad_norm": 0.8846771121025085, + "learning_rate": 4.713991581687827e-05, + "loss": 0.5944, + "num_input_tokens_seen": 2445592, + "step": 9095 + }, + { + "epoch": 2.385321100917431, + "grad_norm": 0.622185468673706, + "learning_rate": 4.7134601395336836e-05, + "loss": 0.399, + "num_input_tokens_seen": 2447256, + "step": 9100 + }, + { + "epoch": 2.386631716906946, + "grad_norm": 0.824977695941925, + "learning_rate": 4.712928234106158e-05, + "loss": 0.3433, + "num_input_tokens_seen": 2448472, + "step": 9105 + }, + { + "epoch": 2.3879423328964613, + "grad_norm": 0.5129672288894653, + "learning_rate": 4.712395865516576e-05, + "loss": 0.435, + "num_input_tokens_seen": 2449880, + "step": 9110 + }, + { + "epoch": 2.3892529488859764, + "grad_norm": 0.867412269115448, + "learning_rate": 4.711863033876362e-05, + "loss": 0.3688, + "num_input_tokens_seen": 2450936, + "step": 9115 + }, + { + "epoch": 2.3905635648754915, + "grad_norm": 0.6785266399383545, + "learning_rate": 4.7113297392970365e-05, + "loss": 0.4681, + "num_input_tokens_seen": 2452488, + "step": 9120 + }, + { + "epoch": 2.3918741808650066, + "grad_norm": 0.999596118927002, + "learning_rate": 4.710795981890218e-05, + "loss": 0.4613, + "num_input_tokens_seen": 2453960, + "step": 9125 + }, + { + "epoch": 2.3931847968545217, + "grad_norm": 0.39308643341064453, + "learning_rate": 4.7102617617676203e-05, + "loss": 0.4686, + "num_input_tokens_seen": 2455208, + "step": 9130 + }, + { + "epoch": 2.3944954128440368, + "grad_norm": 0.749528169631958, + "learning_rate": 4.7097270790410555e-05, + "loss": 0.3053, + "num_input_tokens_seen": 2456152, + "step": 9135 + }, + { + "epoch": 2.395806028833552, + "grad_norm": 0.7013789415359497, + "learning_rate": 4.709191933822431e-05, + "loss": 0.5273, + "num_input_tokens_seen": 2457528, + "step": 9140 + }, + { + "epoch": 2.397116644823067, + "grad_norm": 0.29684191942214966, + "learning_rate": 4.7086563262237514e-05, + "loss": 0.2617, + "num_input_tokens_seen": 2458792, + "step": 9145 + }, + { + "epoch": 2.398427260812582, + "grad_norm": 0.8481683135032654, + "learning_rate": 4.7081202563571194e-05, + "loss": 0.3587, + "num_input_tokens_seen": 2459928, + "step": 9150 + }, + { + "epoch": 2.399737876802097, + "grad_norm": 1.2988947629928589, + "learning_rate": 4.707583724334732e-05, + "loss": 0.4943, + "num_input_tokens_seen": 2461304, + "step": 9155 + }, + { + "epoch": 2.4010484927916123, + "grad_norm": 0.8516278862953186, + "learning_rate": 4.707046730268887e-05, + "loss": 0.5568, + "num_input_tokens_seen": 2462440, + "step": 9160 + }, + { + "epoch": 2.402359108781127, + "grad_norm": 0.5132933855056763, + "learning_rate": 4.706509274271974e-05, + "loss": 0.5643, + "num_input_tokens_seen": 2463864, + "step": 9165 + }, + { + "epoch": 2.4036697247706424, + "grad_norm": 1.810394525527954, + "learning_rate": 4.705971356456483e-05, + "loss": 0.6742, + "num_input_tokens_seen": 2465688, + "step": 9170 + }, + { + "epoch": 2.404980340760157, + "grad_norm": 1.0320502519607544, + "learning_rate": 4.7054329769349984e-05, + "loss": 0.4598, + "num_input_tokens_seen": 2467320, + "step": 9175 + }, + { + "epoch": 2.406290956749672, + "grad_norm": 0.32188284397125244, + "learning_rate": 4.704894135820202e-05, + "loss": 0.3214, + "num_input_tokens_seen": 2468680, + "step": 9180 + }, + { + "epoch": 2.4076015727391873, + "grad_norm": 0.6369513869285583, + "learning_rate": 4.704354833224873e-05, + "loss": 0.6257, + "num_input_tokens_seen": 2470088, + "step": 9185 + }, + { + "epoch": 2.4089121887287024, + "grad_norm": 1.5361237525939941, + "learning_rate": 4.7038150692618864e-05, + "loss": 0.5112, + "num_input_tokens_seen": 2471416, + "step": 9190 + }, + { + "epoch": 2.4102228047182175, + "grad_norm": 0.6292665004730225, + "learning_rate": 4.703274844044214e-05, + "loss": 0.4801, + "num_input_tokens_seen": 2472456, + "step": 9195 + }, + { + "epoch": 2.4115334207077326, + "grad_norm": 1.1214715242385864, + "learning_rate": 4.702734157684924e-05, + "loss": 0.5217, + "num_input_tokens_seen": 2473400, + "step": 9200 + }, + { + "epoch": 2.4128440366972477, + "grad_norm": 0.4324759840965271, + "learning_rate": 4.7021930102971796e-05, + "loss": 0.3236, + "num_input_tokens_seen": 2474616, + "step": 9205 + }, + { + "epoch": 2.4141546526867628, + "grad_norm": 0.8190409541130066, + "learning_rate": 4.701651401994244e-05, + "loss": 0.4183, + "num_input_tokens_seen": 2475752, + "step": 9210 + }, + { + "epoch": 2.415465268676278, + "grad_norm": 0.5602617859840393, + "learning_rate": 4.701109332889474e-05, + "loss": 0.3715, + "num_input_tokens_seen": 2477064, + "step": 9215 + }, + { + "epoch": 2.416775884665793, + "grad_norm": 0.7845823168754578, + "learning_rate": 4.700566803096324e-05, + "loss": 0.4336, + "num_input_tokens_seen": 2478472, + "step": 9220 + }, + { + "epoch": 2.418086500655308, + "grad_norm": 0.37721511721611023, + "learning_rate": 4.700023812728344e-05, + "loss": 0.2746, + "num_input_tokens_seen": 2480408, + "step": 9225 + }, + { + "epoch": 2.419397116644823, + "grad_norm": 0.7825469374656677, + "learning_rate": 4.699480361899181e-05, + "loss": 0.5306, + "num_input_tokens_seen": 2481704, + "step": 9230 + }, + { + "epoch": 2.4207077326343382, + "grad_norm": 0.3516809046268463, + "learning_rate": 4.6989364507225795e-05, + "loss": 0.2422, + "num_input_tokens_seen": 2482776, + "step": 9235 + }, + { + "epoch": 2.4220183486238533, + "grad_norm": 1.525308609008789, + "learning_rate": 4.698392079312378e-05, + "loss": 0.4053, + "num_input_tokens_seen": 2484376, + "step": 9240 + }, + { + "epoch": 2.4233289646133684, + "grad_norm": 0.5903172492980957, + "learning_rate": 4.697847247782513e-05, + "loss": 0.4653, + "num_input_tokens_seen": 2486088, + "step": 9245 + }, + { + "epoch": 2.4246395806028835, + "grad_norm": 0.4734795391559601, + "learning_rate": 4.6973019562470166e-05, + "loss": 0.5628, + "num_input_tokens_seen": 2487320, + "step": 9250 + }, + { + "epoch": 2.4259501965923986, + "grad_norm": 1.045216679573059, + "learning_rate": 4.6967562048200176e-05, + "loss": 0.4731, + "num_input_tokens_seen": 2488424, + "step": 9255 + }, + { + "epoch": 2.4272608125819133, + "grad_norm": 0.898753821849823, + "learning_rate": 4.69620999361574e-05, + "loss": 0.3803, + "num_input_tokens_seen": 2489480, + "step": 9260 + }, + { + "epoch": 2.4285714285714284, + "grad_norm": 0.8720229864120483, + "learning_rate": 4.695663322748506e-05, + "loss": 0.5882, + "num_input_tokens_seen": 2490872, + "step": 9265 + }, + { + "epoch": 2.4298820445609435, + "grad_norm": 1.4414926767349243, + "learning_rate": 4.6951161923327324e-05, + "loss": 0.3992, + "num_input_tokens_seen": 2492152, + "step": 9270 + }, + { + "epoch": 2.4311926605504586, + "grad_norm": 1.4606187343597412, + "learning_rate": 4.694568602482933e-05, + "loss": 0.3816, + "num_input_tokens_seen": 2493640, + "step": 9275 + }, + { + "epoch": 2.4325032765399737, + "grad_norm": 0.4770897924900055, + "learning_rate": 4.694020553313716e-05, + "loss": 0.3267, + "num_input_tokens_seen": 2494792, + "step": 9280 + }, + { + "epoch": 2.4338138925294888, + "grad_norm": 0.599769651889801, + "learning_rate": 4.693472044939788e-05, + "loss": 0.3722, + "num_input_tokens_seen": 2496200, + "step": 9285 + }, + { + "epoch": 2.435124508519004, + "grad_norm": 0.5848119258880615, + "learning_rate": 4.692923077475952e-05, + "loss": 0.4431, + "num_input_tokens_seen": 2497544, + "step": 9290 + }, + { + "epoch": 2.436435124508519, + "grad_norm": 0.6804619431495667, + "learning_rate": 4.6923736510371054e-05, + "loss": 0.5119, + "num_input_tokens_seen": 2498856, + "step": 9295 + }, + { + "epoch": 2.437745740498034, + "grad_norm": 0.6141180396080017, + "learning_rate": 4.69182376573824e-05, + "loss": 0.4525, + "num_input_tokens_seen": 2500136, + "step": 9300 + }, + { + "epoch": 2.439056356487549, + "grad_norm": 1.3354933261871338, + "learning_rate": 4.6912734216944486e-05, + "loss": 0.4249, + "num_input_tokens_seen": 2501304, + "step": 9305 + }, + { + "epoch": 2.4403669724770642, + "grad_norm": 0.7994473576545715, + "learning_rate": 4.6907226190209165e-05, + "loss": 0.4001, + "num_input_tokens_seen": 2502792, + "step": 9310 + }, + { + "epoch": 2.4416775884665793, + "grad_norm": 0.8803660869598389, + "learning_rate": 4.6901713578329255e-05, + "loss": 0.3139, + "num_input_tokens_seen": 2503944, + "step": 9315 + }, + { + "epoch": 2.4429882044560944, + "grad_norm": 0.5943455100059509, + "learning_rate": 4.689619638245853e-05, + "loss": 0.5171, + "num_input_tokens_seen": 2505960, + "step": 9320 + }, + { + "epoch": 2.4442988204456095, + "grad_norm": 1.1517739295959473, + "learning_rate": 4.6890674603751746e-05, + "loss": 0.4452, + "num_input_tokens_seen": 2507480, + "step": 9325 + }, + { + "epoch": 2.4456094364351246, + "grad_norm": 1.131812572479248, + "learning_rate": 4.688514824336459e-05, + "loss": 0.5639, + "num_input_tokens_seen": 2509928, + "step": 9330 + }, + { + "epoch": 2.4469200524246397, + "grad_norm": 1.0003137588500977, + "learning_rate": 4.687961730245371e-05, + "loss": 0.5563, + "num_input_tokens_seen": 2511496, + "step": 9335 + }, + { + "epoch": 2.448230668414155, + "grad_norm": 0.7386566400527954, + "learning_rate": 4.687408178217674e-05, + "loss": 0.3665, + "num_input_tokens_seen": 2512984, + "step": 9340 + }, + { + "epoch": 2.44954128440367, + "grad_norm": 0.5760444402694702, + "learning_rate": 4.6868541683692244e-05, + "loss": 0.4214, + "num_input_tokens_seen": 2514536, + "step": 9345 + }, + { + "epoch": 2.450851900393185, + "grad_norm": 0.6067106127738953, + "learning_rate": 4.6862997008159774e-05, + "loss": 0.3716, + "num_input_tokens_seen": 2515656, + "step": 9350 + }, + { + "epoch": 2.4521625163826997, + "grad_norm": 0.7626911997795105, + "learning_rate": 4.685744775673979e-05, + "loss": 0.5093, + "num_input_tokens_seen": 2517096, + "step": 9355 + }, + { + "epoch": 2.4534731323722148, + "grad_norm": 0.7908127307891846, + "learning_rate": 4.685189393059377e-05, + "loss": 0.5214, + "num_input_tokens_seen": 2518408, + "step": 9360 + }, + { + "epoch": 2.45478374836173, + "grad_norm": 1.7429215908050537, + "learning_rate": 4.68463355308841e-05, + "loss": 0.3649, + "num_input_tokens_seen": 2519512, + "step": 9365 + }, + { + "epoch": 2.456094364351245, + "grad_norm": 0.7082529664039612, + "learning_rate": 4.684077255877415e-05, + "loss": 0.3981, + "num_input_tokens_seen": 2520808, + "step": 9370 + }, + { + "epoch": 2.45740498034076, + "grad_norm": 0.8895736336708069, + "learning_rate": 4.6835205015428246e-05, + "loss": 0.5966, + "num_input_tokens_seen": 2522120, + "step": 9375 + }, + { + "epoch": 2.458715596330275, + "grad_norm": 0.5664262175559998, + "learning_rate": 4.682963290201166e-05, + "loss": 0.4801, + "num_input_tokens_seen": 2523384, + "step": 9380 + }, + { + "epoch": 2.4600262123197902, + "grad_norm": 0.5265970826148987, + "learning_rate": 4.682405621969062e-05, + "loss": 0.4658, + "num_input_tokens_seen": 2525144, + "step": 9385 + }, + { + "epoch": 2.4613368283093053, + "grad_norm": 0.9223127961158752, + "learning_rate": 4.681847496963233e-05, + "loss": 0.4349, + "num_input_tokens_seen": 2526168, + "step": 9390 + }, + { + "epoch": 2.4626474442988204, + "grad_norm": 0.737075686454773, + "learning_rate": 4.6812889153004925e-05, + "loss": 0.6039, + "num_input_tokens_seen": 2527560, + "step": 9395 + }, + { + "epoch": 2.4639580602883355, + "grad_norm": 1.345619559288025, + "learning_rate": 4.6807298770977516e-05, + "loss": 0.4109, + "num_input_tokens_seen": 2529176, + "step": 9400 + }, + { + "epoch": 2.4652686762778506, + "grad_norm": 1.5158886909484863, + "learning_rate": 4.6801703824720146e-05, + "loss": 0.5164, + "num_input_tokens_seen": 2530088, + "step": 9405 + }, + { + "epoch": 2.4665792922673657, + "grad_norm": 0.7851963639259338, + "learning_rate": 4.679610431540384e-05, + "loss": 0.3906, + "num_input_tokens_seen": 2531272, + "step": 9410 + }, + { + "epoch": 2.467889908256881, + "grad_norm": 0.3158690631389618, + "learning_rate": 4.679050024420056e-05, + "loss": 0.5097, + "num_input_tokens_seen": 2532696, + "step": 9415 + }, + { + "epoch": 2.469200524246396, + "grad_norm": 1.4593955278396606, + "learning_rate": 4.678489161228323e-05, + "loss": 0.4403, + "num_input_tokens_seen": 2533640, + "step": 9420 + }, + { + "epoch": 2.470511140235911, + "grad_norm": 0.7721222639083862, + "learning_rate": 4.677927842082572e-05, + "loss": 0.3883, + "num_input_tokens_seen": 2534856, + "step": 9425 + }, + { + "epoch": 2.471821756225426, + "grad_norm": 0.8131278157234192, + "learning_rate": 4.6773660671002885e-05, + "loss": 0.3576, + "num_input_tokens_seen": 2536152, + "step": 9430 + }, + { + "epoch": 2.473132372214941, + "grad_norm": 0.4710918068885803, + "learning_rate": 4.676803836399048e-05, + "loss": 0.5674, + "num_input_tokens_seen": 2537560, + "step": 9435 + }, + { + "epoch": 2.474442988204456, + "grad_norm": 0.5571103096008301, + "learning_rate": 4.6762411500965265e-05, + "loss": 0.3399, + "num_input_tokens_seen": 2538792, + "step": 9440 + }, + { + "epoch": 2.4757536041939714, + "grad_norm": 0.6823135614395142, + "learning_rate": 4.675678008310492e-05, + "loss": 0.5611, + "num_input_tokens_seen": 2539896, + "step": 9445 + }, + { + "epoch": 2.477064220183486, + "grad_norm": 0.8835307359695435, + "learning_rate": 4.6751144111588094e-05, + "loss": 0.4399, + "num_input_tokens_seen": 2541032, + "step": 9450 + }, + { + "epoch": 2.478374836173001, + "grad_norm": 0.797244668006897, + "learning_rate": 4.6745503587594396e-05, + "loss": 0.3441, + "num_input_tokens_seen": 2542120, + "step": 9455 + }, + { + "epoch": 2.4796854521625162, + "grad_norm": 0.34900546073913574, + "learning_rate": 4.6739858512304366e-05, + "loss": 0.5922, + "num_input_tokens_seen": 2543096, + "step": 9460 + }, + { + "epoch": 2.4809960681520313, + "grad_norm": 0.8782224059104919, + "learning_rate": 4.673420888689952e-05, + "loss": 0.3405, + "num_input_tokens_seen": 2544328, + "step": 9465 + }, + { + "epoch": 2.4823066841415464, + "grad_norm": 1.9879024028778076, + "learning_rate": 4.67285547125623e-05, + "loss": 0.4576, + "num_input_tokens_seen": 2545448, + "step": 9470 + }, + { + "epoch": 2.4836173001310615, + "grad_norm": 1.0291030406951904, + "learning_rate": 4.6722895990476126e-05, + "loss": 0.4073, + "num_input_tokens_seen": 2546584, + "step": 9475 + }, + { + "epoch": 2.4849279161205766, + "grad_norm": 1.3115874528884888, + "learning_rate": 4.671723272182535e-05, + "loss": 0.3813, + "num_input_tokens_seen": 2547832, + "step": 9480 + }, + { + "epoch": 2.4862385321100917, + "grad_norm": 1.3681095838546753, + "learning_rate": 4.67115649077953e-05, + "loss": 0.5927, + "num_input_tokens_seen": 2549016, + "step": 9485 + }, + { + "epoch": 2.487549148099607, + "grad_norm": 0.7394666075706482, + "learning_rate": 4.670589254957222e-05, + "loss": 0.3309, + "num_input_tokens_seen": 2550392, + "step": 9490 + }, + { + "epoch": 2.488859764089122, + "grad_norm": 0.611640453338623, + "learning_rate": 4.6700215648343336e-05, + "loss": 0.3627, + "num_input_tokens_seen": 2551992, + "step": 9495 + }, + { + "epoch": 2.490170380078637, + "grad_norm": 0.4381481111049652, + "learning_rate": 4.669453420529682e-05, + "loss": 0.3466, + "num_input_tokens_seen": 2553736, + "step": 9500 + }, + { + "epoch": 2.491480996068152, + "grad_norm": 1.8333654403686523, + "learning_rate": 4.668884822162177e-05, + "loss": 0.503, + "num_input_tokens_seen": 2554968, + "step": 9505 + }, + { + "epoch": 2.492791612057667, + "grad_norm": 0.7153487205505371, + "learning_rate": 4.668315769850827e-05, + "loss": 0.3246, + "num_input_tokens_seen": 2556120, + "step": 9510 + }, + { + "epoch": 2.4941022280471823, + "grad_norm": 0.591378927230835, + "learning_rate": 4.667746263714733e-05, + "loss": 0.4275, + "num_input_tokens_seen": 2557560, + "step": 9515 + }, + { + "epoch": 2.4954128440366974, + "grad_norm": 0.7974472045898438, + "learning_rate": 4.667176303873091e-05, + "loss": 0.4757, + "num_input_tokens_seen": 2558888, + "step": 9520 + }, + { + "epoch": 2.4967234600262125, + "grad_norm": 0.5759813189506531, + "learning_rate": 4.666605890445193e-05, + "loss": 0.3123, + "num_input_tokens_seen": 2560024, + "step": 9525 + }, + { + "epoch": 2.4980340760157276, + "grad_norm": 1.594193696975708, + "learning_rate": 4.6660350235504265e-05, + "loss": 0.4953, + "num_input_tokens_seen": 2561192, + "step": 9530 + }, + { + "epoch": 2.4993446920052422, + "grad_norm": 0.39447373151779175, + "learning_rate": 4.6654637033082715e-05, + "loss": 0.4147, + "num_input_tokens_seen": 2562648, + "step": 9535 + }, + { + "epoch": 2.5006553079947578, + "grad_norm": 0.6383748054504395, + "learning_rate": 4.6648919298383055e-05, + "loss": 0.6815, + "num_input_tokens_seen": 2563624, + "step": 9540 + }, + { + "epoch": 2.5006553079947578, + "eval_loss": 0.5050681829452515, + "eval_runtime": 16.8632, + "eval_samples_per_second": 50.287, + "eval_steps_per_second": 25.144, + "num_input_tokens_seen": 2563624, + "step": 9540 + }, + { + "epoch": 2.5019659239842724, + "grad_norm": 0.5586893558502197, + "learning_rate": 4.664319703260199e-05, + "loss": 0.5692, + "num_input_tokens_seen": 2565160, + "step": 9545 + }, + { + "epoch": 2.5032765399737875, + "grad_norm": 0.6282629370689392, + "learning_rate": 4.6637470236937186e-05, + "loss": 0.5569, + "num_input_tokens_seen": 2566232, + "step": 9550 + }, + { + "epoch": 2.5045871559633026, + "grad_norm": 0.8085957169532776, + "learning_rate": 4.663173891258724e-05, + "loss": 0.2908, + "num_input_tokens_seen": 2567400, + "step": 9555 + }, + { + "epoch": 2.5058977719528177, + "grad_norm": 0.4027717709541321, + "learning_rate": 4.662600306075172e-05, + "loss": 0.2949, + "num_input_tokens_seen": 2568840, + "step": 9560 + }, + { + "epoch": 2.507208387942333, + "grad_norm": 0.6091422438621521, + "learning_rate": 4.662026268263113e-05, + "loss": 0.6695, + "num_input_tokens_seen": 2570392, + "step": 9565 + }, + { + "epoch": 2.508519003931848, + "grad_norm": 0.7206957340240479, + "learning_rate": 4.6614517779426917e-05, + "loss": 0.5657, + "num_input_tokens_seen": 2572072, + "step": 9570 + }, + { + "epoch": 2.509829619921363, + "grad_norm": 1.2050092220306396, + "learning_rate": 4.660876835234148e-05, + "loss": 0.6318, + "num_input_tokens_seen": 2573096, + "step": 9575 + }, + { + "epoch": 2.511140235910878, + "grad_norm": 2.078015089035034, + "learning_rate": 4.660301440257816e-05, + "loss": 0.5966, + "num_input_tokens_seen": 2574424, + "step": 9580 + }, + { + "epoch": 2.512450851900393, + "grad_norm": 1.2848365306854248, + "learning_rate": 4.659725593134125e-05, + "loss": 0.3285, + "num_input_tokens_seen": 2575688, + "step": 9585 + }, + { + "epoch": 2.5137614678899083, + "grad_norm": 0.8038976788520813, + "learning_rate": 4.6591492939836e-05, + "loss": 0.5266, + "num_input_tokens_seen": 2577240, + "step": 9590 + }, + { + "epoch": 2.5150720838794234, + "grad_norm": 1.061181664466858, + "learning_rate": 4.6585725429268585e-05, + "loss": 0.5294, + "num_input_tokens_seen": 2578760, + "step": 9595 + }, + { + "epoch": 2.5163826998689385, + "grad_norm": 0.4464094340801239, + "learning_rate": 4.657995340084613e-05, + "loss": 0.4552, + "num_input_tokens_seen": 2580328, + "step": 9600 + }, + { + "epoch": 2.5176933158584536, + "grad_norm": 1.0025874376296997, + "learning_rate": 4.657417685577671e-05, + "loss": 0.5462, + "num_input_tokens_seen": 2581400, + "step": 9605 + }, + { + "epoch": 2.5190039318479687, + "grad_norm": 0.5359145998954773, + "learning_rate": 4.656839579526936e-05, + "loss": 0.4296, + "num_input_tokens_seen": 2582568, + "step": 9610 + }, + { + "epoch": 2.5203145478374838, + "grad_norm": 1.9621425867080688, + "learning_rate": 4.6562610220534035e-05, + "loss": 0.4198, + "num_input_tokens_seen": 2583816, + "step": 9615 + }, + { + "epoch": 2.5216251638269984, + "grad_norm": 0.8270935416221619, + "learning_rate": 4.6556820132781646e-05, + "loss": 0.359, + "num_input_tokens_seen": 2585128, + "step": 9620 + }, + { + "epoch": 2.522935779816514, + "grad_norm": 0.872007429599762, + "learning_rate": 4.6551025533224056e-05, + "loss": 0.3314, + "num_input_tokens_seen": 2586120, + "step": 9625 + }, + { + "epoch": 2.5242463958060286, + "grad_norm": 1.3936868906021118, + "learning_rate": 4.654522642307405e-05, + "loss": 0.4678, + "num_input_tokens_seen": 2587352, + "step": 9630 + }, + { + "epoch": 2.525557011795544, + "grad_norm": 0.404256671667099, + "learning_rate": 4.653942280354539e-05, + "loss": 0.4005, + "num_input_tokens_seen": 2588840, + "step": 9635 + }, + { + "epoch": 2.526867627785059, + "grad_norm": 0.9405596256256104, + "learning_rate": 4.6533614675852755e-05, + "loss": 0.3503, + "num_input_tokens_seen": 2590232, + "step": 9640 + }, + { + "epoch": 2.528178243774574, + "grad_norm": 1.7424206733703613, + "learning_rate": 4.652780204121177e-05, + "loss": 0.5377, + "num_input_tokens_seen": 2591400, + "step": 9645 + }, + { + "epoch": 2.529488859764089, + "grad_norm": 0.6380326747894287, + "learning_rate": 4.652198490083901e-05, + "loss": 0.374, + "num_input_tokens_seen": 2592488, + "step": 9650 + }, + { + "epoch": 2.530799475753604, + "grad_norm": 0.5175490379333496, + "learning_rate": 4.6516163255952004e-05, + "loss": 0.3477, + "num_input_tokens_seen": 2593544, + "step": 9655 + }, + { + "epoch": 2.532110091743119, + "grad_norm": 0.9549685716629028, + "learning_rate": 4.651033710776921e-05, + "loss": 0.3644, + "num_input_tokens_seen": 2594696, + "step": 9660 + }, + { + "epoch": 2.5334207077326343, + "grad_norm": 2.5089681148529053, + "learning_rate": 4.6504506457510025e-05, + "loss": 0.5314, + "num_input_tokens_seen": 2595864, + "step": 9665 + }, + { + "epoch": 2.5347313237221494, + "grad_norm": 0.913062334060669, + "learning_rate": 4.649867130639479e-05, + "loss": 0.6784, + "num_input_tokens_seen": 2597080, + "step": 9670 + }, + { + "epoch": 2.5360419397116645, + "grad_norm": 0.8141002058982849, + "learning_rate": 4.64928316556448e-05, + "loss": 0.3739, + "num_input_tokens_seen": 2598200, + "step": 9675 + }, + { + "epoch": 2.5373525557011796, + "grad_norm": 0.8186720013618469, + "learning_rate": 4.648698750648228e-05, + "loss": 0.4584, + "num_input_tokens_seen": 2599512, + "step": 9680 + }, + { + "epoch": 2.5386631716906947, + "grad_norm": 0.5895242094993591, + "learning_rate": 4.64811388601304e-05, + "loss": 0.4413, + "num_input_tokens_seen": 2600984, + "step": 9685 + }, + { + "epoch": 2.5399737876802098, + "grad_norm": 1.3597151041030884, + "learning_rate": 4.647528571781328e-05, + "loss": 0.4793, + "num_input_tokens_seen": 2602024, + "step": 9690 + }, + { + "epoch": 2.541284403669725, + "grad_norm": 0.7782910466194153, + "learning_rate": 4.646942808075596e-05, + "loss": 0.6566, + "num_input_tokens_seen": 2603848, + "step": 9695 + }, + { + "epoch": 2.54259501965924, + "grad_norm": 0.5546464323997498, + "learning_rate": 4.646356595018444e-05, + "loss": 0.3064, + "num_input_tokens_seen": 2605384, + "step": 9700 + }, + { + "epoch": 2.543905635648755, + "grad_norm": 0.9527125358581543, + "learning_rate": 4.6457699327325655e-05, + "loss": 0.5232, + "num_input_tokens_seen": 2606600, + "step": 9705 + }, + { + "epoch": 2.54521625163827, + "grad_norm": 0.7811403274536133, + "learning_rate": 4.645182821340747e-05, + "loss": 0.4464, + "num_input_tokens_seen": 2608072, + "step": 9710 + }, + { + "epoch": 2.546526867627785, + "grad_norm": 0.6205427050590515, + "learning_rate": 4.644595260965872e-05, + "loss": 0.344, + "num_input_tokens_seen": 2609400, + "step": 9715 + }, + { + "epoch": 2.5478374836173003, + "grad_norm": 0.6719418168067932, + "learning_rate": 4.644007251730913e-05, + "loss": 0.5349, + "num_input_tokens_seen": 2610648, + "step": 9720 + }, + { + "epoch": 2.549148099606815, + "grad_norm": 0.471405953168869, + "learning_rate": 4.643418793758941e-05, + "loss": 0.4658, + "num_input_tokens_seen": 2612392, + "step": 9725 + }, + { + "epoch": 2.5504587155963305, + "grad_norm": 0.7170855402946472, + "learning_rate": 4.642829887173119e-05, + "loss": 0.4238, + "num_input_tokens_seen": 2614088, + "step": 9730 + }, + { + "epoch": 2.551769331585845, + "grad_norm": 0.617766261100769, + "learning_rate": 4.642240532096705e-05, + "loss": 0.3854, + "num_input_tokens_seen": 2615256, + "step": 9735 + }, + { + "epoch": 2.5530799475753603, + "grad_norm": 2.2041208744049072, + "learning_rate": 4.641650728653049e-05, + "loss": 0.5256, + "num_input_tokens_seen": 2616568, + "step": 9740 + }, + { + "epoch": 2.5543905635648754, + "grad_norm": 0.4194200336933136, + "learning_rate": 4.641060476965595e-05, + "loss": 0.2882, + "num_input_tokens_seen": 2617896, + "step": 9745 + }, + { + "epoch": 2.5557011795543905, + "grad_norm": 0.6202077865600586, + "learning_rate": 4.640469777157883e-05, + "loss": 0.4395, + "num_input_tokens_seen": 2618968, + "step": 9750 + }, + { + "epoch": 2.5570117955439056, + "grad_norm": 1.1112138032913208, + "learning_rate": 4.6398786293535455e-05, + "loss": 0.6892, + "num_input_tokens_seen": 2619928, + "step": 9755 + }, + { + "epoch": 2.5583224115334207, + "grad_norm": 0.6673231720924377, + "learning_rate": 4.639287033676308e-05, + "loss": 0.4248, + "num_input_tokens_seen": 2621048, + "step": 9760 + }, + { + "epoch": 2.5596330275229358, + "grad_norm": 0.7857370972633362, + "learning_rate": 4.638694990249991e-05, + "loss": 0.4194, + "num_input_tokens_seen": 2622152, + "step": 9765 + }, + { + "epoch": 2.560943643512451, + "grad_norm": 0.516413152217865, + "learning_rate": 4.638102499198508e-05, + "loss": 0.439, + "num_input_tokens_seen": 2624104, + "step": 9770 + }, + { + "epoch": 2.562254259501966, + "grad_norm": 0.4197818338871002, + "learning_rate": 4.637509560645865e-05, + "loss": 0.4915, + "num_input_tokens_seen": 2625416, + "step": 9775 + }, + { + "epoch": 2.563564875491481, + "grad_norm": 0.5935743451118469, + "learning_rate": 4.636916174716166e-05, + "loss": 0.7944, + "num_input_tokens_seen": 2626600, + "step": 9780 + }, + { + "epoch": 2.564875491480996, + "grad_norm": 0.6731524467468262, + "learning_rate": 4.636322341533603e-05, + "loss": 0.5528, + "num_input_tokens_seen": 2627864, + "step": 9785 + }, + { + "epoch": 2.5661861074705112, + "grad_norm": 0.5784062147140503, + "learning_rate": 4.6357280612224656e-05, + "loss": 0.7865, + "num_input_tokens_seen": 2628984, + "step": 9790 + }, + { + "epoch": 2.5674967234600263, + "grad_norm": 0.6891485452651978, + "learning_rate": 4.635133333907135e-05, + "loss": 0.4205, + "num_input_tokens_seen": 2630376, + "step": 9795 + }, + { + "epoch": 2.5688073394495414, + "grad_norm": 1.488141655921936, + "learning_rate": 4.634538159712088e-05, + "loss": 0.2772, + "num_input_tokens_seen": 2631448, + "step": 9800 + }, + { + "epoch": 2.5701179554390565, + "grad_norm": 0.6945971846580505, + "learning_rate": 4.6339425387618906e-05, + "loss": 0.329, + "num_input_tokens_seen": 2632664, + "step": 9805 + }, + { + "epoch": 2.571428571428571, + "grad_norm": 0.5930336117744446, + "learning_rate": 4.633346471181207e-05, + "loss": 0.4213, + "num_input_tokens_seen": 2633896, + "step": 9810 + }, + { + "epoch": 2.5727391874180867, + "grad_norm": 0.7883399128913879, + "learning_rate": 4.632749957094794e-05, + "loss": 0.4111, + "num_input_tokens_seen": 2635416, + "step": 9815 + }, + { + "epoch": 2.5740498034076014, + "grad_norm": 0.9578903913497925, + "learning_rate": 4.6321529966275e-05, + "loss": 0.4024, + "num_input_tokens_seen": 2636632, + "step": 9820 + }, + { + "epoch": 2.575360419397117, + "grad_norm": 0.7661036252975464, + "learning_rate": 4.6315555899042686e-05, + "loss": 0.6242, + "num_input_tokens_seen": 2637672, + "step": 9825 + }, + { + "epoch": 2.5766710353866316, + "grad_norm": 0.8521307110786438, + "learning_rate": 4.630957737050134e-05, + "loss": 0.5375, + "num_input_tokens_seen": 2639064, + "step": 9830 + }, + { + "epoch": 2.5779816513761467, + "grad_norm": 1.1476099491119385, + "learning_rate": 4.630359438190227e-05, + "loss": 0.4342, + "num_input_tokens_seen": 2640184, + "step": 9835 + }, + { + "epoch": 2.5792922673656618, + "grad_norm": 0.9575284123420715, + "learning_rate": 4.629760693449772e-05, + "loss": 0.4967, + "num_input_tokens_seen": 2641400, + "step": 9840 + }, + { + "epoch": 2.580602883355177, + "grad_norm": 0.46713072061538696, + "learning_rate": 4.6291615029540826e-05, + "loss": 0.4355, + "num_input_tokens_seen": 2642792, + "step": 9845 + }, + { + "epoch": 2.581913499344692, + "grad_norm": 0.8191307783126831, + "learning_rate": 4.62856186682857e-05, + "loss": 0.4068, + "num_input_tokens_seen": 2643992, + "step": 9850 + }, + { + "epoch": 2.583224115334207, + "grad_norm": 1.5879559516906738, + "learning_rate": 4.627961785198736e-05, + "loss": 0.4328, + "num_input_tokens_seen": 2644808, + "step": 9855 + }, + { + "epoch": 2.584534731323722, + "grad_norm": 1.37531578540802, + "learning_rate": 4.627361258190178e-05, + "loss": 0.4022, + "num_input_tokens_seen": 2646168, + "step": 9860 + }, + { + "epoch": 2.5858453473132372, + "grad_norm": 0.8207103610038757, + "learning_rate": 4.626760285928584e-05, + "loss": 0.4511, + "num_input_tokens_seen": 2647512, + "step": 9865 + }, + { + "epoch": 2.5871559633027523, + "grad_norm": 0.9335265159606934, + "learning_rate": 4.626158868539737e-05, + "loss": 0.2714, + "num_input_tokens_seen": 2648584, + "step": 9870 + }, + { + "epoch": 2.5884665792922674, + "grad_norm": 2.637413501739502, + "learning_rate": 4.625557006149512e-05, + "loss": 0.4856, + "num_input_tokens_seen": 2650120, + "step": 9875 + }, + { + "epoch": 2.5897771952817825, + "grad_norm": 0.501270055770874, + "learning_rate": 4.624954698883879e-05, + "loss": 0.581, + "num_input_tokens_seen": 2651624, + "step": 9880 + }, + { + "epoch": 2.5910878112712976, + "grad_norm": 0.8653784990310669, + "learning_rate": 4.624351946868899e-05, + "loss": 0.5313, + "num_input_tokens_seen": 2652888, + "step": 9885 + }, + { + "epoch": 2.5923984272608127, + "grad_norm": 1.8621772527694702, + "learning_rate": 4.6237487502307265e-05, + "loss": 0.4677, + "num_input_tokens_seen": 2654184, + "step": 9890 + }, + { + "epoch": 2.593709043250328, + "grad_norm": 1.1377614736557007, + "learning_rate": 4.623145109095611e-05, + "loss": 0.3713, + "num_input_tokens_seen": 2655832, + "step": 9895 + }, + { + "epoch": 2.595019659239843, + "grad_norm": 1.4108810424804688, + "learning_rate": 4.622541023589893e-05, + "loss": 0.6268, + "num_input_tokens_seen": 2657288, + "step": 9900 + }, + { + "epoch": 2.5963302752293576, + "grad_norm": 0.7277272343635559, + "learning_rate": 4.621936493840005e-05, + "loss": 0.6626, + "num_input_tokens_seen": 2658760, + "step": 9905 + }, + { + "epoch": 2.597640891218873, + "grad_norm": 2.6802361011505127, + "learning_rate": 4.621331519972476e-05, + "loss": 0.7849, + "num_input_tokens_seen": 2659928, + "step": 9910 + }, + { + "epoch": 2.5989515072083877, + "grad_norm": 0.5163223743438721, + "learning_rate": 4.6207261021139246e-05, + "loss": 0.3692, + "num_input_tokens_seen": 2661160, + "step": 9915 + }, + { + "epoch": 2.6002621231979033, + "grad_norm": 1.190356969833374, + "learning_rate": 4.620120240391065e-05, + "loss": 0.4008, + "num_input_tokens_seen": 2662328, + "step": 9920 + }, + { + "epoch": 2.601572739187418, + "grad_norm": 0.698269248008728, + "learning_rate": 4.6195139349307024e-05, + "loss": 0.3099, + "num_input_tokens_seen": 2663672, + "step": 9925 + }, + { + "epoch": 2.602883355176933, + "grad_norm": 0.7145283222198486, + "learning_rate": 4.6189071858597355e-05, + "loss": 0.4407, + "num_input_tokens_seen": 2665000, + "step": 9930 + }, + { + "epoch": 2.604193971166448, + "grad_norm": 0.6114339232444763, + "learning_rate": 4.6182999933051554e-05, + "loss": 0.3869, + "num_input_tokens_seen": 2666632, + "step": 9935 + }, + { + "epoch": 2.6055045871559632, + "grad_norm": 0.6427066922187805, + "learning_rate": 4.617692357394047e-05, + "loss": 0.3716, + "num_input_tokens_seen": 2668392, + "step": 9940 + }, + { + "epoch": 2.6068152031454783, + "grad_norm": 1.0437637567520142, + "learning_rate": 4.617084278253587e-05, + "loss": 0.3868, + "num_input_tokens_seen": 2669736, + "step": 9945 + }, + { + "epoch": 2.6081258191349934, + "grad_norm": 0.532774806022644, + "learning_rate": 4.616475756011046e-05, + "loss": 0.508, + "num_input_tokens_seen": 2671112, + "step": 9950 + }, + { + "epoch": 2.6094364351245085, + "grad_norm": 0.485359787940979, + "learning_rate": 4.615866790793786e-05, + "loss": 0.4306, + "num_input_tokens_seen": 2672856, + "step": 9955 + }, + { + "epoch": 2.6107470511140236, + "grad_norm": 0.6307424306869507, + "learning_rate": 4.6152573827292636e-05, + "loss": 0.5392, + "num_input_tokens_seen": 2674008, + "step": 9960 + }, + { + "epoch": 2.6120576671035387, + "grad_norm": 0.6106417179107666, + "learning_rate": 4.6146475319450244e-05, + "loss": 0.7406, + "num_input_tokens_seen": 2675176, + "step": 9965 + }, + { + "epoch": 2.613368283093054, + "grad_norm": 0.962934136390686, + "learning_rate": 4.6140372385687114e-05, + "loss": 0.4467, + "num_input_tokens_seen": 2676664, + "step": 9970 + }, + { + "epoch": 2.614678899082569, + "grad_norm": 0.7228918671607971, + "learning_rate": 4.6134265027280574e-05, + "loss": 0.5381, + "num_input_tokens_seen": 2678104, + "step": 9975 + }, + { + "epoch": 2.615989515072084, + "grad_norm": 1.3031198978424072, + "learning_rate": 4.6128153245508876e-05, + "loss": 0.471, + "num_input_tokens_seen": 2679112, + "step": 9980 + }, + { + "epoch": 2.617300131061599, + "grad_norm": 0.826630175113678, + "learning_rate": 4.612203704165121e-05, + "loss": 0.3432, + "num_input_tokens_seen": 2680600, + "step": 9985 + }, + { + "epoch": 2.618610747051114, + "grad_norm": 1.2472847700119019, + "learning_rate": 4.6115916416987686e-05, + "loss": 0.6246, + "num_input_tokens_seen": 2681672, + "step": 9990 + }, + { + "epoch": 2.6199213630406293, + "grad_norm": 1.5826998949050903, + "learning_rate": 4.610979137279935e-05, + "loss": 0.324, + "num_input_tokens_seen": 2682792, + "step": 9995 + }, + { + "epoch": 2.621231979030144, + "grad_norm": 2.1137051582336426, + "learning_rate": 4.6103661910368146e-05, + "loss": 0.4565, + "num_input_tokens_seen": 2684088, + "step": 10000 + }, + { + "epoch": 2.6225425950196595, + "grad_norm": 0.5440008640289307, + "learning_rate": 4.6097528030976964e-05, + "loss": 0.2685, + "num_input_tokens_seen": 2685624, + "step": 10005 + }, + { + "epoch": 2.623853211009174, + "grad_norm": 0.4231520891189575, + "learning_rate": 4.609138973590963e-05, + "loss": 0.5173, + "num_input_tokens_seen": 2686808, + "step": 10010 + }, + { + "epoch": 2.625163826998689, + "grad_norm": 0.4156995415687561, + "learning_rate": 4.6085247026450866e-05, + "loss": 0.4452, + "num_input_tokens_seen": 2687976, + "step": 10015 + }, + { + "epoch": 2.6264744429882043, + "grad_norm": 1.236219882965088, + "learning_rate": 4.607909990388632e-05, + "loss": 0.506, + "num_input_tokens_seen": 2689400, + "step": 10020 + }, + { + "epoch": 2.6277850589777194, + "grad_norm": 0.7271056771278381, + "learning_rate": 4.607294836950261e-05, + "loss": 0.435, + "num_input_tokens_seen": 2690664, + "step": 10025 + }, + { + "epoch": 2.6290956749672345, + "grad_norm": 0.5120954513549805, + "learning_rate": 4.606679242458719e-05, + "loss": 0.4291, + "num_input_tokens_seen": 2691848, + "step": 10030 + }, + { + "epoch": 2.6304062909567496, + "grad_norm": 0.5757666826248169, + "learning_rate": 4.606063207042853e-05, + "loss": 0.3754, + "num_input_tokens_seen": 2693128, + "step": 10035 + }, + { + "epoch": 2.6317169069462647, + "grad_norm": 0.8303540349006653, + "learning_rate": 4.6054467308315975e-05, + "loss": 0.4434, + "num_input_tokens_seen": 2694392, + "step": 10040 + }, + { + "epoch": 2.63302752293578, + "grad_norm": 0.7658789157867432, + "learning_rate": 4.604829813953978e-05, + "loss": 0.5752, + "num_input_tokens_seen": 2695560, + "step": 10045 + }, + { + "epoch": 2.634338138925295, + "grad_norm": 0.9482907056808472, + "learning_rate": 4.6042124565391155e-05, + "loss": 0.4206, + "num_input_tokens_seen": 2696872, + "step": 10050 + }, + { + "epoch": 2.63564875491481, + "grad_norm": 3.441767454147339, + "learning_rate": 4.603594658716221e-05, + "loss": 0.8657, + "num_input_tokens_seen": 2698104, + "step": 10055 + }, + { + "epoch": 2.636959370904325, + "grad_norm": 0.6376848220825195, + "learning_rate": 4.6029764206146006e-05, + "loss": 0.2645, + "num_input_tokens_seen": 2699800, + "step": 10060 + }, + { + "epoch": 2.63826998689384, + "grad_norm": 0.7433394193649292, + "learning_rate": 4.602357742363649e-05, + "loss": 0.2466, + "num_input_tokens_seen": 2701400, + "step": 10065 + }, + { + "epoch": 2.6395806028833553, + "grad_norm": 0.4109465479850769, + "learning_rate": 4.601738624092853e-05, + "loss": 0.4039, + "num_input_tokens_seen": 2702744, + "step": 10070 + }, + { + "epoch": 2.6408912188728704, + "grad_norm": 0.7813175916671753, + "learning_rate": 4.601119065931796e-05, + "loss": 0.2803, + "num_input_tokens_seen": 2704296, + "step": 10075 + }, + { + "epoch": 2.6422018348623855, + "grad_norm": 1.003705620765686, + "learning_rate": 4.600499068010148e-05, + "loss": 0.484, + "num_input_tokens_seen": 2705992, + "step": 10080 + }, + { + "epoch": 2.6435124508519, + "grad_norm": 0.682684063911438, + "learning_rate": 4.599878630457674e-05, + "loss": 0.4983, + "num_input_tokens_seen": 2707816, + "step": 10085 + }, + { + "epoch": 2.6448230668414157, + "grad_norm": 1.1148085594177246, + "learning_rate": 4.599257753404233e-05, + "loss": 0.5145, + "num_input_tokens_seen": 2709144, + "step": 10090 + }, + { + "epoch": 2.6461336828309303, + "grad_norm": 1.133147120475769, + "learning_rate": 4.59863643697977e-05, + "loss": 0.5396, + "num_input_tokens_seen": 2710152, + "step": 10095 + }, + { + "epoch": 2.647444298820446, + "grad_norm": 1.3006136417388916, + "learning_rate": 4.5980146813143264e-05, + "loss": 0.6607, + "num_input_tokens_seen": 2711304, + "step": 10100 + }, + { + "epoch": 2.6487549148099605, + "grad_norm": 1.441888451576233, + "learning_rate": 4.597392486538036e-05, + "loss": 0.5502, + "num_input_tokens_seen": 2712328, + "step": 10105 + }, + { + "epoch": 2.6500655307994756, + "grad_norm": 0.6324690580368042, + "learning_rate": 4.596769852781122e-05, + "loss": 0.5064, + "num_input_tokens_seen": 2714120, + "step": 10110 + }, + { + "epoch": 2.6513761467889907, + "grad_norm": 0.6596447825431824, + "learning_rate": 4.596146780173901e-05, + "loss": 0.5362, + "num_input_tokens_seen": 2715800, + "step": 10115 + }, + { + "epoch": 2.652686762778506, + "grad_norm": 0.8150302171707153, + "learning_rate": 4.59552326884678e-05, + "loss": 0.5368, + "num_input_tokens_seen": 2717016, + "step": 10120 + }, + { + "epoch": 2.653997378768021, + "grad_norm": 0.08114083111286163, + "learning_rate": 4.5948993189302605e-05, + "loss": 0.6015, + "num_input_tokens_seen": 2719176, + "step": 10125 + }, + { + "epoch": 2.655307994757536, + "grad_norm": 0.5264469981193542, + "learning_rate": 4.5942749305549325e-05, + "loss": 0.3876, + "num_input_tokens_seen": 2720456, + "step": 10130 + }, + { + "epoch": 2.656618610747051, + "grad_norm": 0.6302986741065979, + "learning_rate": 4.593650103851481e-05, + "loss": 0.3088, + "num_input_tokens_seen": 2721576, + "step": 10135 + }, + { + "epoch": 2.657929226736566, + "grad_norm": 2.7403454780578613, + "learning_rate": 4.5930248389506794e-05, + "loss": 0.7981, + "num_input_tokens_seen": 2722680, + "step": 10140 + }, + { + "epoch": 2.6592398427260813, + "grad_norm": 0.6101958155632019, + "learning_rate": 4.592399135983396e-05, + "loss": 0.3838, + "num_input_tokens_seen": 2724104, + "step": 10145 + }, + { + "epoch": 2.6605504587155964, + "grad_norm": 1.0048497915267944, + "learning_rate": 4.591772995080589e-05, + "loss": 0.3664, + "num_input_tokens_seen": 2725432, + "step": 10150 + }, + { + "epoch": 2.6618610747051115, + "grad_norm": 0.44213300943374634, + "learning_rate": 4.591146416373308e-05, + "loss": 0.5431, + "num_input_tokens_seen": 2726712, + "step": 10155 + }, + { + "epoch": 2.6631716906946266, + "grad_norm": 0.5014444589614868, + "learning_rate": 4.590519399992695e-05, + "loss": 0.4842, + "num_input_tokens_seen": 2727880, + "step": 10160 + }, + { + "epoch": 2.6644823066841417, + "grad_norm": 1.0315272808074951, + "learning_rate": 4.5898919460699845e-05, + "loss": 0.4633, + "num_input_tokens_seen": 2729096, + "step": 10165 + }, + { + "epoch": 2.6657929226736568, + "grad_norm": 0.7555910348892212, + "learning_rate": 4.589264054736501e-05, + "loss": 0.4841, + "num_input_tokens_seen": 2730264, + "step": 10170 + }, + { + "epoch": 2.667103538663172, + "grad_norm": 0.7580534815788269, + "learning_rate": 4.58863572612366e-05, + "loss": 0.3895, + "num_input_tokens_seen": 2731352, + "step": 10175 + }, + { + "epoch": 2.6684141546526865, + "grad_norm": 0.7978902459144592, + "learning_rate": 4.5880069603629705e-05, + "loss": 0.4891, + "num_input_tokens_seen": 2732552, + "step": 10180 + }, + { + "epoch": 2.669724770642202, + "grad_norm": 1.798952579498291, + "learning_rate": 4.587377757586032e-05, + "loss": 0.516, + "num_input_tokens_seen": 2734232, + "step": 10185 + }, + { + "epoch": 2.6710353866317167, + "grad_norm": 0.6879321932792664, + "learning_rate": 4.586748117924536e-05, + "loss": 0.4562, + "num_input_tokens_seen": 2735480, + "step": 10190 + }, + { + "epoch": 2.6723460026212322, + "grad_norm": 0.7035673260688782, + "learning_rate": 4.586118041510264e-05, + "loss": 0.5471, + "num_input_tokens_seen": 2736824, + "step": 10195 + }, + { + "epoch": 2.673656618610747, + "grad_norm": 1.8364315032958984, + "learning_rate": 4.585487528475091e-05, + "loss": 0.3381, + "num_input_tokens_seen": 2737992, + "step": 10200 + }, + { + "epoch": 2.674967234600262, + "grad_norm": 0.9650053977966309, + "learning_rate": 4.584856578950981e-05, + "loss": 0.4401, + "num_input_tokens_seen": 2739272, + "step": 10205 + }, + { + "epoch": 2.676277850589777, + "grad_norm": 0.510420560836792, + "learning_rate": 4.584225193069992e-05, + "loss": 0.4005, + "num_input_tokens_seen": 2740840, + "step": 10210 + }, + { + "epoch": 2.677588466579292, + "grad_norm": 0.6798452138900757, + "learning_rate": 4.583593370964271e-05, + "loss": 0.4715, + "num_input_tokens_seen": 2742632, + "step": 10215 + }, + { + "epoch": 2.6788990825688073, + "grad_norm": 1.073808193206787, + "learning_rate": 4.582961112766058e-05, + "loss": 0.4595, + "num_input_tokens_seen": 2743944, + "step": 10220 + }, + { + "epoch": 2.6802096985583224, + "grad_norm": 1.0556143522262573, + "learning_rate": 4.582328418607682e-05, + "loss": 0.4414, + "num_input_tokens_seen": 2745064, + "step": 10225 + }, + { + "epoch": 2.6815203145478375, + "grad_norm": 1.03921639919281, + "learning_rate": 4.581695288621568e-05, + "loss": 0.5779, + "num_input_tokens_seen": 2746312, + "step": 10230 + }, + { + "epoch": 2.6828309305373526, + "grad_norm": 0.618941605091095, + "learning_rate": 4.581061722940225e-05, + "loss": 0.4849, + "num_input_tokens_seen": 2747672, + "step": 10235 + }, + { + "epoch": 2.6841415465268676, + "grad_norm": 0.7894508838653564, + "learning_rate": 4.580427721696261e-05, + "loss": 0.5557, + "num_input_tokens_seen": 2748680, + "step": 10240 + }, + { + "epoch": 2.6854521625163827, + "grad_norm": 1.4883856773376465, + "learning_rate": 4.579793285022368e-05, + "loss": 0.3893, + "num_input_tokens_seen": 2749944, + "step": 10245 + }, + { + "epoch": 2.686762778505898, + "grad_norm": 0.44883492588996887, + "learning_rate": 4.579158413051335e-05, + "loss": 0.4983, + "num_input_tokens_seen": 2751320, + "step": 10250 + }, + { + "epoch": 2.688073394495413, + "grad_norm": 0.5216087698936462, + "learning_rate": 4.578523105916038e-05, + "loss": 0.3491, + "num_input_tokens_seen": 2753064, + "step": 10255 + }, + { + "epoch": 2.689384010484928, + "grad_norm": 0.7345361709594727, + "learning_rate": 4.577887363749447e-05, + "loss": 0.6213, + "num_input_tokens_seen": 2754024, + "step": 10260 + }, + { + "epoch": 2.690694626474443, + "grad_norm": 0.4912348985671997, + "learning_rate": 4.57725118668462e-05, + "loss": 0.6495, + "num_input_tokens_seen": 2755800, + "step": 10265 + }, + { + "epoch": 2.6920052424639582, + "grad_norm": 0.5804017186164856, + "learning_rate": 4.5766145748547104e-05, + "loss": 0.3653, + "num_input_tokens_seen": 2756872, + "step": 10270 + }, + { + "epoch": 2.693315858453473, + "grad_norm": 0.710513710975647, + "learning_rate": 4.575977528392957e-05, + "loss": 0.4177, + "num_input_tokens_seen": 2758440, + "step": 10275 + }, + { + "epoch": 2.6946264744429884, + "grad_norm": 0.4878494143486023, + "learning_rate": 4.5753400474326946e-05, + "loss": 0.3065, + "num_input_tokens_seen": 2759816, + "step": 10280 + }, + { + "epoch": 2.695937090432503, + "grad_norm": 0.6477996706962585, + "learning_rate": 4.574702132107346e-05, + "loss": 0.5045, + "num_input_tokens_seen": 2761400, + "step": 10285 + }, + { + "epoch": 2.6972477064220186, + "grad_norm": 0.5125251412391663, + "learning_rate": 4.5740637825504265e-05, + "loss": 0.4377, + "num_input_tokens_seen": 2762392, + "step": 10290 + }, + { + "epoch": 2.6985583224115333, + "grad_norm": 1.055795669555664, + "learning_rate": 4.5734249988955405e-05, + "loss": 0.3996, + "num_input_tokens_seen": 2763864, + "step": 10295 + }, + { + "epoch": 2.6998689384010484, + "grad_norm": 0.9898115396499634, + "learning_rate": 4.572785781276385e-05, + "loss": 0.4463, + "num_input_tokens_seen": 2764856, + "step": 10300 + }, + { + "epoch": 2.7011795543905635, + "grad_norm": 2.3240602016448975, + "learning_rate": 4.5721461298267466e-05, + "loss": 0.5552, + "num_input_tokens_seen": 2765944, + "step": 10305 + }, + { + "epoch": 2.7024901703800785, + "grad_norm": 1.234783411026001, + "learning_rate": 4.5715060446805044e-05, + "loss": 0.3079, + "num_input_tokens_seen": 2766888, + "step": 10310 + }, + { + "epoch": 2.7038007863695936, + "grad_norm": 0.5504478216171265, + "learning_rate": 4.570865525971626e-05, + "loss": 0.3607, + "num_input_tokens_seen": 2768328, + "step": 10315 + }, + { + "epoch": 2.7051114023591087, + "grad_norm": 0.9003111720085144, + "learning_rate": 4.570224573834171e-05, + "loss": 0.3476, + "num_input_tokens_seen": 2769256, + "step": 10320 + }, + { + "epoch": 2.706422018348624, + "grad_norm": 1.3159343004226685, + "learning_rate": 4.569583188402289e-05, + "loss": 0.4212, + "num_input_tokens_seen": 2770744, + "step": 10325 + }, + { + "epoch": 2.707732634338139, + "grad_norm": 0.9117443561553955, + "learning_rate": 4.5689413698102226e-05, + "loss": 0.4677, + "num_input_tokens_seen": 2772200, + "step": 10330 + }, + { + "epoch": 2.709043250327654, + "grad_norm": 0.44229206442832947, + "learning_rate": 4.5682991181923024e-05, + "loss": 0.3446, + "num_input_tokens_seen": 2773752, + "step": 10335 + }, + { + "epoch": 2.710353866317169, + "grad_norm": 1.4026477336883545, + "learning_rate": 4.56765643368295e-05, + "loss": 0.5616, + "num_input_tokens_seen": 2774952, + "step": 10340 + }, + { + "epoch": 2.711664482306684, + "grad_norm": 1.5174744129180908, + "learning_rate": 4.5670133164166795e-05, + "loss": 0.3965, + "num_input_tokens_seen": 2776072, + "step": 10345 + }, + { + "epoch": 2.7129750982961993, + "grad_norm": 0.4308152496814728, + "learning_rate": 4.566369766528093e-05, + "loss": 0.3625, + "num_input_tokens_seen": 2777560, + "step": 10350 + }, + { + "epoch": 2.7142857142857144, + "grad_norm": 1.4060852527618408, + "learning_rate": 4.565725784151884e-05, + "loss": 0.581, + "num_input_tokens_seen": 2778456, + "step": 10355 + }, + { + "epoch": 2.7155963302752295, + "grad_norm": 1.0076477527618408, + "learning_rate": 4.565081369422839e-05, + "loss": 0.4191, + "num_input_tokens_seen": 2779672, + "step": 10360 + }, + { + "epoch": 2.7169069462647446, + "grad_norm": 2.0669167041778564, + "learning_rate": 4.56443652247583e-05, + "loss": 0.3994, + "num_input_tokens_seen": 2780824, + "step": 10365 + }, + { + "epoch": 2.7182175622542593, + "grad_norm": 1.5388480424880981, + "learning_rate": 4.5637912434458254e-05, + "loss": 0.944, + "num_input_tokens_seen": 2781960, + "step": 10370 + }, + { + "epoch": 2.719528178243775, + "grad_norm": 0.9293700456619263, + "learning_rate": 4.563145532467878e-05, + "loss": 0.4834, + "num_input_tokens_seen": 2783064, + "step": 10375 + }, + { + "epoch": 2.7208387942332894, + "grad_norm": 2.3961849212646484, + "learning_rate": 4.562499389677137e-05, + "loss": 0.612, + "num_input_tokens_seen": 2784280, + "step": 10380 + }, + { + "epoch": 2.722149410222805, + "grad_norm": 1.1853164434432983, + "learning_rate": 4.561852815208837e-05, + "loss": 0.7265, + "num_input_tokens_seen": 2785736, + "step": 10385 + }, + { + "epoch": 2.7234600262123196, + "grad_norm": 0.6867725849151611, + "learning_rate": 4.561205809198305e-05, + "loss": 0.4172, + "num_input_tokens_seen": 2787064, + "step": 10390 + }, + { + "epoch": 2.7247706422018347, + "grad_norm": 0.9144600033760071, + "learning_rate": 4.560558371780959e-05, + "loss": 0.482, + "num_input_tokens_seen": 2788120, + "step": 10395 + }, + { + "epoch": 2.72608125819135, + "grad_norm": 2.607327699661255, + "learning_rate": 4.559910503092306e-05, + "loss": 0.4257, + "num_input_tokens_seen": 2789288, + "step": 10400 + }, + { + "epoch": 2.727391874180865, + "grad_norm": 0.7760504484176636, + "learning_rate": 4.559262203267945e-05, + "loss": 0.5317, + "num_input_tokens_seen": 2790280, + "step": 10405 + }, + { + "epoch": 2.72870249017038, + "grad_norm": 0.6977182030677795, + "learning_rate": 4.5586134724435614e-05, + "loss": 0.265, + "num_input_tokens_seen": 2791320, + "step": 10410 + }, + { + "epoch": 2.730013106159895, + "grad_norm": 0.8163205981254578, + "learning_rate": 4.557964310754935e-05, + "loss": 0.4466, + "num_input_tokens_seen": 2792856, + "step": 10415 + }, + { + "epoch": 2.73132372214941, + "grad_norm": 0.5699107646942139, + "learning_rate": 4.557314718337935e-05, + "loss": 0.4862, + "num_input_tokens_seen": 2794200, + "step": 10420 + }, + { + "epoch": 2.7326343381389253, + "grad_norm": 0.780342161655426, + "learning_rate": 4.556664695328519e-05, + "loss": 0.3957, + "num_input_tokens_seen": 2795192, + "step": 10425 + }, + { + "epoch": 2.7339449541284404, + "grad_norm": 0.733879566192627, + "learning_rate": 4.556014241862736e-05, + "loss": 0.3202, + "num_input_tokens_seen": 2796296, + "step": 10430 + }, + { + "epoch": 2.7352555701179555, + "grad_norm": 0.6407439708709717, + "learning_rate": 4.555363358076725e-05, + "loss": 0.3949, + "num_input_tokens_seen": 2797512, + "step": 10435 + }, + { + "epoch": 2.7365661861074706, + "grad_norm": 3.851332664489746, + "learning_rate": 4.5547120441067136e-05, + "loss": 0.5276, + "num_input_tokens_seen": 2798504, + "step": 10440 + }, + { + "epoch": 2.7378768020969857, + "grad_norm": 0.47543397545814514, + "learning_rate": 4.554060300089022e-05, + "loss": 0.3932, + "num_input_tokens_seen": 2799832, + "step": 10445 + }, + { + "epoch": 2.739187418086501, + "grad_norm": 1.0656700134277344, + "learning_rate": 4.553408126160059e-05, + "loss": 0.3827, + "num_input_tokens_seen": 2801048, + "step": 10450 + }, + { + "epoch": 2.740498034076016, + "grad_norm": 0.6428584456443787, + "learning_rate": 4.552755522456324e-05, + "loss": 0.4713, + "num_input_tokens_seen": 2802568, + "step": 10455 + }, + { + "epoch": 2.741808650065531, + "grad_norm": 1.2680127620697021, + "learning_rate": 4.552102489114404e-05, + "loss": 0.5176, + "num_input_tokens_seen": 2804184, + "step": 10460 + }, + { + "epoch": 2.7431192660550456, + "grad_norm": 0.6635900735855103, + "learning_rate": 4.551449026270979e-05, + "loss": 0.3866, + "num_input_tokens_seen": 2805512, + "step": 10465 + }, + { + "epoch": 2.744429882044561, + "grad_norm": 1.3592543601989746, + "learning_rate": 4.550795134062817e-05, + "loss": 0.5029, + "num_input_tokens_seen": 2806664, + "step": 10470 + }, + { + "epoch": 2.745740498034076, + "grad_norm": 0.7272982597351074, + "learning_rate": 4.5501408126267784e-05, + "loss": 0.4511, + "num_input_tokens_seen": 2808200, + "step": 10475 + }, + { + "epoch": 2.747051114023591, + "grad_norm": 0.7710971832275391, + "learning_rate": 4.5494860620998086e-05, + "loss": 0.3884, + "num_input_tokens_seen": 2809528, + "step": 10480 + }, + { + "epoch": 2.748361730013106, + "grad_norm": 2.5225064754486084, + "learning_rate": 4.548830882618948e-05, + "loss": 0.5136, + "num_input_tokens_seen": 2810984, + "step": 10485 + }, + { + "epoch": 2.749672346002621, + "grad_norm": 0.8898077607154846, + "learning_rate": 4.5481752743213234e-05, + "loss": 0.3959, + "num_input_tokens_seen": 2812600, + "step": 10490 + }, + { + "epoch": 2.750982961992136, + "grad_norm": 1.1412345170974731, + "learning_rate": 4.5475192373441525e-05, + "loss": 0.4235, + "num_input_tokens_seen": 2814568, + "step": 10495 + }, + { + "epoch": 2.7522935779816513, + "grad_norm": 0.6587729454040527, + "learning_rate": 4.5468627718247436e-05, + "loss": 0.6947, + "num_input_tokens_seen": 2816216, + "step": 10500 + }, + { + "epoch": 2.7536041939711664, + "grad_norm": 1.1748895645141602, + "learning_rate": 4.546205877900494e-05, + "loss": 0.521, + "num_input_tokens_seen": 2817560, + "step": 10505 + }, + { + "epoch": 2.7549148099606815, + "grad_norm": 1.6601595878601074, + "learning_rate": 4.5455485557088885e-05, + "loss": 0.5897, + "num_input_tokens_seen": 2818616, + "step": 10510 + }, + { + "epoch": 2.7562254259501966, + "grad_norm": 0.8550077080726624, + "learning_rate": 4.544890805387505e-05, + "loss": 0.4069, + "num_input_tokens_seen": 2819976, + "step": 10515 + }, + { + "epoch": 2.7575360419397117, + "grad_norm": 0.37487030029296875, + "learning_rate": 4.544232627074009e-05, + "loss": 0.4999, + "num_input_tokens_seen": 2821464, + "step": 10520 + }, + { + "epoch": 2.758846657929227, + "grad_norm": 0.5795673131942749, + "learning_rate": 4.5435740209061574e-05, + "loss": 0.4374, + "num_input_tokens_seen": 2822744, + "step": 10525 + }, + { + "epoch": 2.760157273918742, + "grad_norm": 0.8453917503356934, + "learning_rate": 4.5429149870217936e-05, + "loss": 0.7152, + "num_input_tokens_seen": 2823992, + "step": 10530 + }, + { + "epoch": 2.761467889908257, + "grad_norm": 0.909865140914917, + "learning_rate": 4.5422555255588524e-05, + "loss": 0.3585, + "num_input_tokens_seen": 2825256, + "step": 10535 + }, + { + "epoch": 2.762778505897772, + "grad_norm": 1.0576485395431519, + "learning_rate": 4.541595636655359e-05, + "loss": 0.4345, + "num_input_tokens_seen": 2826312, + "step": 10540 + }, + { + "epoch": 2.764089121887287, + "grad_norm": 1.10515296459198, + "learning_rate": 4.540935320449426e-05, + "loss": 0.3674, + "num_input_tokens_seen": 2827816, + "step": 10545 + }, + { + "epoch": 2.765399737876802, + "grad_norm": 0.9496889114379883, + "learning_rate": 4.540274577079258e-05, + "loss": 0.552, + "num_input_tokens_seen": 2829272, + "step": 10550 + }, + { + "epoch": 2.7667103538663174, + "grad_norm": 0.7884969115257263, + "learning_rate": 4.539613406683145e-05, + "loss": 0.5816, + "num_input_tokens_seen": 2830648, + "step": 10555 + }, + { + "epoch": 2.768020969855832, + "grad_norm": 0.9200642108917236, + "learning_rate": 4.538951809399471e-05, + "loss": 0.6404, + "num_input_tokens_seen": 2832088, + "step": 10560 + }, + { + "epoch": 2.7693315858453476, + "grad_norm": 0.5102277994155884, + "learning_rate": 4.538289785366706e-05, + "loss": 0.3271, + "num_input_tokens_seen": 2833720, + "step": 10565 + }, + { + "epoch": 2.770642201834862, + "grad_norm": 0.7396383285522461, + "learning_rate": 4.53762733472341e-05, + "loss": 0.5498, + "num_input_tokens_seen": 2834888, + "step": 10570 + }, + { + "epoch": 2.7719528178243773, + "grad_norm": 0.8610491156578064, + "learning_rate": 4.536964457608235e-05, + "loss": 0.3686, + "num_input_tokens_seen": 2836232, + "step": 10575 + }, + { + "epoch": 2.7732634338138924, + "grad_norm": 1.4054054021835327, + "learning_rate": 4.5363011541599185e-05, + "loss": 0.5454, + "num_input_tokens_seen": 2837720, + "step": 10580 + }, + { + "epoch": 2.7745740498034075, + "grad_norm": 1.1228190660476685, + "learning_rate": 4.535637424517288e-05, + "loss": 0.589, + "num_input_tokens_seen": 2839000, + "step": 10585 + }, + { + "epoch": 2.7758846657929226, + "grad_norm": 1.054721474647522, + "learning_rate": 4.5349732688192624e-05, + "loss": 0.4453, + "num_input_tokens_seen": 2840360, + "step": 10590 + }, + { + "epoch": 2.7771952817824377, + "grad_norm": 0.9682180881500244, + "learning_rate": 4.534308687204848e-05, + "loss": 0.4681, + "num_input_tokens_seen": 2841576, + "step": 10595 + }, + { + "epoch": 2.778505897771953, + "grad_norm": 1.4915987253189087, + "learning_rate": 4.53364367981314e-05, + "loss": 0.5985, + "num_input_tokens_seen": 2842888, + "step": 10600 + }, + { + "epoch": 2.779816513761468, + "grad_norm": 2.034644365310669, + "learning_rate": 4.5329782467833236e-05, + "loss": 0.328, + "num_input_tokens_seen": 2844040, + "step": 10605 + }, + { + "epoch": 2.781127129750983, + "grad_norm": 3.6704931259155273, + "learning_rate": 4.532312388254674e-05, + "loss": 0.4162, + "num_input_tokens_seen": 2845208, + "step": 10610 + }, + { + "epoch": 2.782437745740498, + "grad_norm": 0.332381010055542, + "learning_rate": 4.5316461043665516e-05, + "loss": 0.5642, + "num_input_tokens_seen": 2846792, + "step": 10615 + }, + { + "epoch": 2.783748361730013, + "grad_norm": 0.7294674515724182, + "learning_rate": 4.53097939525841e-05, + "loss": 0.3334, + "num_input_tokens_seen": 2848120, + "step": 10620 + }, + { + "epoch": 2.7850589777195283, + "grad_norm": 0.4438440799713135, + "learning_rate": 4.530312261069791e-05, + "loss": 0.4653, + "num_input_tokens_seen": 2849544, + "step": 10625 + }, + { + "epoch": 2.7863695937090434, + "grad_norm": 0.8253045678138733, + "learning_rate": 4.5296447019403245e-05, + "loss": 0.5307, + "num_input_tokens_seen": 2850648, + "step": 10630 + }, + { + "epoch": 2.7876802096985585, + "grad_norm": 1.3716167211532593, + "learning_rate": 4.5289767180097275e-05, + "loss": 0.5779, + "num_input_tokens_seen": 2851864, + "step": 10635 + }, + { + "epoch": 2.7889908256880735, + "grad_norm": 1.5017629861831665, + "learning_rate": 4.528308309417809e-05, + "loss": 0.3657, + "num_input_tokens_seen": 2852984, + "step": 10640 + }, + { + "epoch": 2.790301441677588, + "grad_norm": 0.592571496963501, + "learning_rate": 4.527639476304467e-05, + "loss": 0.4504, + "num_input_tokens_seen": 2854152, + "step": 10645 + }, + { + "epoch": 2.7916120576671037, + "grad_norm": 0.5654451251029968, + "learning_rate": 4.526970218809686e-05, + "loss": 0.3793, + "num_input_tokens_seen": 2856104, + "step": 10650 + }, + { + "epoch": 2.7929226736566184, + "grad_norm": 1.35752272605896, + "learning_rate": 4.52630053707354e-05, + "loss": 0.5374, + "num_input_tokens_seen": 2857176, + "step": 10655 + }, + { + "epoch": 2.794233289646134, + "grad_norm": 0.7907735109329224, + "learning_rate": 4.525630431236193e-05, + "loss": 0.3588, + "num_input_tokens_seen": 2858632, + "step": 10660 + }, + { + "epoch": 2.7955439056356486, + "grad_norm": 0.5908024311065674, + "learning_rate": 4.524959901437897e-05, + "loss": 0.4613, + "num_input_tokens_seen": 2860312, + "step": 10665 + }, + { + "epoch": 2.7968545216251637, + "grad_norm": 3.4110724925994873, + "learning_rate": 4.5242889478189926e-05, + "loss": 0.4443, + "num_input_tokens_seen": 2861512, + "step": 10670 + }, + { + "epoch": 2.7981651376146788, + "grad_norm": 0.9561628699302673, + "learning_rate": 4.5236175705199094e-05, + "loss": 0.4729, + "num_input_tokens_seen": 2862904, + "step": 10675 + }, + { + "epoch": 2.799475753604194, + "grad_norm": 0.9576904773712158, + "learning_rate": 4.522945769681164e-05, + "loss": 0.4538, + "num_input_tokens_seen": 2864232, + "step": 10680 + }, + { + "epoch": 2.800786369593709, + "grad_norm": 5.3556437492370605, + "learning_rate": 4.522273545443365e-05, + "loss": 0.2642, + "num_input_tokens_seen": 2865608, + "step": 10685 + }, + { + "epoch": 2.802096985583224, + "grad_norm": 0.9686478972434998, + "learning_rate": 4.521600897947208e-05, + "loss": 0.4864, + "num_input_tokens_seen": 2867224, + "step": 10690 + }, + { + "epoch": 2.803407601572739, + "grad_norm": 0.5533281564712524, + "learning_rate": 4.520927827333475e-05, + "loss": 0.4801, + "num_input_tokens_seen": 2868344, + "step": 10695 + }, + { + "epoch": 2.8047182175622543, + "grad_norm": 0.8369457125663757, + "learning_rate": 4.5202543337430405e-05, + "loss": 0.3929, + "num_input_tokens_seen": 2869528, + "step": 10700 + }, + { + "epoch": 2.8060288335517694, + "grad_norm": 0.41548311710357666, + "learning_rate": 4.519580417316863e-05, + "loss": 0.2381, + "num_input_tokens_seen": 2871112, + "step": 10705 + }, + { + "epoch": 2.8073394495412844, + "grad_norm": 0.8489300012588501, + "learning_rate": 4.518906078195996e-05, + "loss": 0.4471, + "num_input_tokens_seen": 2872184, + "step": 10710 + }, + { + "epoch": 2.8086500655307995, + "grad_norm": 0.9253180027008057, + "learning_rate": 4.518231316521574e-05, + "loss": 0.4297, + "num_input_tokens_seen": 2873448, + "step": 10715 + }, + { + "epoch": 2.8099606815203146, + "grad_norm": 0.7073834538459778, + "learning_rate": 4.517556132434825e-05, + "loss": 0.3554, + "num_input_tokens_seen": 2874664, + "step": 10720 + }, + { + "epoch": 2.8112712975098297, + "grad_norm": 0.8615631461143494, + "learning_rate": 4.5168805260770633e-05, + "loss": 0.499, + "num_input_tokens_seen": 2876072, + "step": 10725 + }, + { + "epoch": 2.812581913499345, + "grad_norm": 1.0766059160232544, + "learning_rate": 4.516204497589692e-05, + "loss": 0.4057, + "num_input_tokens_seen": 2877144, + "step": 10730 + }, + { + "epoch": 2.81389252948886, + "grad_norm": 0.6943012475967407, + "learning_rate": 4.5155280471142036e-05, + "loss": 0.4213, + "num_input_tokens_seen": 2878680, + "step": 10735 + }, + { + "epoch": 2.8152031454783746, + "grad_norm": 0.49379628896713257, + "learning_rate": 4.514851174792177e-05, + "loss": 0.4944, + "num_input_tokens_seen": 2879640, + "step": 10740 + }, + { + "epoch": 2.81651376146789, + "grad_norm": 0.8957775235176086, + "learning_rate": 4.514173880765282e-05, + "loss": 0.5437, + "num_input_tokens_seen": 2880904, + "step": 10745 + }, + { + "epoch": 2.8178243774574048, + "grad_norm": 1.1308798789978027, + "learning_rate": 4.513496165175273e-05, + "loss": 0.5003, + "num_input_tokens_seen": 2882312, + "step": 10750 + }, + { + "epoch": 2.8191349934469203, + "grad_norm": 0.9781889319419861, + "learning_rate": 4.512818028163996e-05, + "loss": 0.361, + "num_input_tokens_seen": 2883624, + "step": 10755 + }, + { + "epoch": 2.820445609436435, + "grad_norm": 0.604550838470459, + "learning_rate": 4.512139469873384e-05, + "loss": 0.3979, + "num_input_tokens_seen": 2884680, + "step": 10760 + }, + { + "epoch": 2.82175622542595, + "grad_norm": 3.1519813537597656, + "learning_rate": 4.511460490445457e-05, + "loss": 0.3262, + "num_input_tokens_seen": 2885800, + "step": 10765 + }, + { + "epoch": 2.823066841415465, + "grad_norm": 0.24502448737621307, + "learning_rate": 4.510781090022325e-05, + "loss": 0.3678, + "num_input_tokens_seen": 2887352, + "step": 10770 + }, + { + "epoch": 2.8243774574049803, + "grad_norm": 2.0142836570739746, + "learning_rate": 4.5101012687461854e-05, + "loss": 0.4346, + "num_input_tokens_seen": 2888632, + "step": 10775 + }, + { + "epoch": 2.8256880733944953, + "grad_norm": 1.0801759958267212, + "learning_rate": 4.509421026759323e-05, + "loss": 0.3862, + "num_input_tokens_seen": 2889576, + "step": 10780 + }, + { + "epoch": 2.8269986893840104, + "grad_norm": 1.209962010383606, + "learning_rate": 4.508740364204113e-05, + "loss": 0.3761, + "num_input_tokens_seen": 2891320, + "step": 10785 + }, + { + "epoch": 2.8283093053735255, + "grad_norm": 0.6540942192077637, + "learning_rate": 4.508059281223015e-05, + "loss": 0.4267, + "num_input_tokens_seen": 2892984, + "step": 10790 + }, + { + "epoch": 2.8296199213630406, + "grad_norm": 1.9557862281799316, + "learning_rate": 4.507377777958578e-05, + "loss": 0.4885, + "num_input_tokens_seen": 2894232, + "step": 10795 + }, + { + "epoch": 2.8309305373525557, + "grad_norm": 1.4571800231933594, + "learning_rate": 4.5066958545534415e-05, + "loss": 0.3749, + "num_input_tokens_seen": 2895224, + "step": 10800 + }, + { + "epoch": 2.832241153342071, + "grad_norm": 0.48643797636032104, + "learning_rate": 4.50601351115033e-05, + "loss": 0.4454, + "num_input_tokens_seen": 2896536, + "step": 10805 + }, + { + "epoch": 2.833551769331586, + "grad_norm": 0.4705289900302887, + "learning_rate": 4.5053307478920564e-05, + "loss": 0.4718, + "num_input_tokens_seen": 2897992, + "step": 10810 + }, + { + "epoch": 2.834862385321101, + "grad_norm": 0.5485686659812927, + "learning_rate": 4.504647564921523e-05, + "loss": 0.4173, + "num_input_tokens_seen": 2899496, + "step": 10815 + }, + { + "epoch": 2.836173001310616, + "grad_norm": 6.219308376312256, + "learning_rate": 4.503963962381717e-05, + "loss": 0.4792, + "num_input_tokens_seen": 2900696, + "step": 10820 + }, + { + "epoch": 2.837483617300131, + "grad_norm": 0.8857677578926086, + "learning_rate": 4.503279940415717e-05, + "loss": 0.3424, + "num_input_tokens_seen": 2902072, + "step": 10825 + }, + { + "epoch": 2.8387942332896463, + "grad_norm": 1.0540648698806763, + "learning_rate": 4.5025954991666865e-05, + "loss": 0.4219, + "num_input_tokens_seen": 2903256, + "step": 10830 + }, + { + "epoch": 2.840104849279161, + "grad_norm": 0.7026126384735107, + "learning_rate": 4.501910638777878e-05, + "loss": 0.4682, + "num_input_tokens_seen": 2904632, + "step": 10835 + }, + { + "epoch": 2.8414154652686765, + "grad_norm": 0.4173215925693512, + "learning_rate": 4.5012253593926324e-05, + "loss": 0.3468, + "num_input_tokens_seen": 2905848, + "step": 10840 + }, + { + "epoch": 2.842726081258191, + "grad_norm": 1.0808206796646118, + "learning_rate": 4.5005396611543765e-05, + "loss": 0.3642, + "num_input_tokens_seen": 2907208, + "step": 10845 + }, + { + "epoch": 2.8440366972477067, + "grad_norm": 0.49489977955818176, + "learning_rate": 4.499853544206626e-05, + "loss": 0.441, + "num_input_tokens_seen": 2908424, + "step": 10850 + }, + { + "epoch": 2.8453473132372213, + "grad_norm": 0.700878381729126, + "learning_rate": 4.499167008692985e-05, + "loss": 0.5106, + "num_input_tokens_seen": 2909496, + "step": 10855 + }, + { + "epoch": 2.8466579292267364, + "grad_norm": 0.4980717897415161, + "learning_rate": 4.4984800547571425e-05, + "loss": 0.3329, + "num_input_tokens_seen": 2911144, + "step": 10860 + }, + { + "epoch": 2.8479685452162515, + "grad_norm": 0.4602383077144623, + "learning_rate": 4.497792682542878e-05, + "loss": 0.359, + "num_input_tokens_seen": 2912536, + "step": 10865 + }, + { + "epoch": 2.8492791612057666, + "grad_norm": 1.2245653867721558, + "learning_rate": 4.4971048921940575e-05, + "loss": 0.4173, + "num_input_tokens_seen": 2913672, + "step": 10870 + }, + { + "epoch": 2.8505897771952817, + "grad_norm": 2.491593360900879, + "learning_rate": 4.496416683854633e-05, + "loss": 0.5925, + "num_input_tokens_seen": 2914616, + "step": 10875 + }, + { + "epoch": 2.851900393184797, + "grad_norm": 0.7316901683807373, + "learning_rate": 4.495728057668647e-05, + "loss": 0.5661, + "num_input_tokens_seen": 2915912, + "step": 10880 + }, + { + "epoch": 2.853211009174312, + "grad_norm": 0.7494410276412964, + "learning_rate": 4.4950390137802254e-05, + "loss": 0.5452, + "num_input_tokens_seen": 2917336, + "step": 10885 + }, + { + "epoch": 2.854521625163827, + "grad_norm": 1.1941134929656982, + "learning_rate": 4.4943495523335865e-05, + "loss": 0.6085, + "num_input_tokens_seen": 2918776, + "step": 10890 + }, + { + "epoch": 2.855832241153342, + "grad_norm": 0.8500843048095703, + "learning_rate": 4.493659673473032e-05, + "loss": 0.4389, + "num_input_tokens_seen": 2921144, + "step": 10895 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.6089376211166382, + "learning_rate": 4.492969377342952e-05, + "loss": 0.2832, + "num_input_tokens_seen": 2922776, + "step": 10900 + }, + { + "epoch": 2.8584534731323723, + "grad_norm": 0.8657322525978088, + "learning_rate": 4.4922786640878265e-05, + "loss": 0.4084, + "num_input_tokens_seen": 2923992, + "step": 10905 + }, + { + "epoch": 2.8597640891218874, + "grad_norm": 0.745597779750824, + "learning_rate": 4.491587533852218e-05, + "loss": 0.3447, + "num_input_tokens_seen": 2925432, + "step": 10910 + }, + { + "epoch": 2.8610747051114025, + "grad_norm": 0.789451003074646, + "learning_rate": 4.49089598678078e-05, + "loss": 0.4495, + "num_input_tokens_seen": 2926776, + "step": 10915 + }, + { + "epoch": 2.8623853211009176, + "grad_norm": 0.9795705676078796, + "learning_rate": 4.490204023018252e-05, + "loss": 0.4077, + "num_input_tokens_seen": 2927656, + "step": 10920 + }, + { + "epoch": 2.8636959370904327, + "grad_norm": 0.43577131628990173, + "learning_rate": 4.4895116427094605e-05, + "loss": 0.5122, + "num_input_tokens_seen": 2928808, + "step": 10925 + }, + { + "epoch": 2.8650065530799473, + "grad_norm": 1.49210524559021, + "learning_rate": 4.48881884599932e-05, + "loss": 0.4024, + "num_input_tokens_seen": 2930472, + "step": 10930 + }, + { + "epoch": 2.866317169069463, + "grad_norm": 1.2760343551635742, + "learning_rate": 4.4881256330328314e-05, + "loss": 0.3917, + "num_input_tokens_seen": 2931976, + "step": 10935 + }, + { + "epoch": 2.8676277850589775, + "grad_norm": 0.7538425326347351, + "learning_rate": 4.4874320039550834e-05, + "loss": 0.4214, + "num_input_tokens_seen": 2932968, + "step": 10940 + }, + { + "epoch": 2.8689384010484926, + "grad_norm": 1.194776177406311, + "learning_rate": 4.48673795891125e-05, + "loss": 0.4417, + "num_input_tokens_seen": 2934072, + "step": 10945 + }, + { + "epoch": 2.8702490170380077, + "grad_norm": 2.296032667160034, + "learning_rate": 4.4860434980465954e-05, + "loss": 0.5831, + "num_input_tokens_seen": 2935560, + "step": 10950 + }, + { + "epoch": 2.871559633027523, + "grad_norm": 0.7445668578147888, + "learning_rate": 4.485348621506468e-05, + "loss": 0.3504, + "num_input_tokens_seen": 2937112, + "step": 10955 + }, + { + "epoch": 2.872870249017038, + "grad_norm": 1.4359325170516968, + "learning_rate": 4.484653329436305e-05, + "loss": 0.4683, + "num_input_tokens_seen": 2938360, + "step": 10960 + }, + { + "epoch": 2.874180865006553, + "grad_norm": 0.8545170426368713, + "learning_rate": 4.483957621981629e-05, + "loss": 0.5955, + "num_input_tokens_seen": 2939624, + "step": 10965 + }, + { + "epoch": 2.875491480996068, + "grad_norm": 0.6143866777420044, + "learning_rate": 4.483261499288051e-05, + "loss": 0.3089, + "num_input_tokens_seen": 2941032, + "step": 10970 + }, + { + "epoch": 2.876802096985583, + "grad_norm": 1.1661524772644043, + "learning_rate": 4.482564961501268e-05, + "loss": 0.719, + "num_input_tokens_seen": 2942200, + "step": 10975 + }, + { + "epoch": 2.8781127129750983, + "grad_norm": 1.388861894607544, + "learning_rate": 4.481868008767063e-05, + "loss": 0.3612, + "num_input_tokens_seen": 2945208, + "step": 10980 + }, + { + "epoch": 2.8794233289646134, + "grad_norm": 1.2847976684570312, + "learning_rate": 4.4811706412313094e-05, + "loss": 0.4045, + "num_input_tokens_seen": 2946248, + "step": 10985 + }, + { + "epoch": 2.8807339449541285, + "grad_norm": 0.4639476239681244, + "learning_rate": 4.4804728590399635e-05, + "loss": 0.4159, + "num_input_tokens_seen": 2948088, + "step": 10990 + }, + { + "epoch": 2.8820445609436436, + "grad_norm": 1.0761128664016724, + "learning_rate": 4.47977466233907e-05, + "loss": 0.3017, + "num_input_tokens_seen": 2949240, + "step": 10995 + }, + { + "epoch": 2.8833551769331587, + "grad_norm": 2.6987836360931396, + "learning_rate": 4.479076051274761e-05, + "loss": 0.5508, + "num_input_tokens_seen": 2950328, + "step": 11000 + }, + { + "epoch": 2.8846657929226738, + "grad_norm": 0.36604344844818115, + "learning_rate": 4.478377025993254e-05, + "loss": 0.362, + "num_input_tokens_seen": 2951848, + "step": 11005 + }, + { + "epoch": 2.885976408912189, + "grad_norm": 1.454829216003418, + "learning_rate": 4.477677586640854e-05, + "loss": 0.4997, + "num_input_tokens_seen": 2952888, + "step": 11010 + }, + { + "epoch": 2.8872870249017035, + "grad_norm": 1.1379426717758179, + "learning_rate": 4.476977733363951e-05, + "loss": 0.5136, + "num_input_tokens_seen": 2953912, + "step": 11015 + }, + { + "epoch": 2.888597640891219, + "grad_norm": 1.7827322483062744, + "learning_rate": 4.4762774663090256e-05, + "loss": 0.4842, + "num_input_tokens_seen": 2955288, + "step": 11020 + }, + { + "epoch": 2.8899082568807337, + "grad_norm": 0.6971645951271057, + "learning_rate": 4.4755767856226405e-05, + "loss": 0.34, + "num_input_tokens_seen": 2956536, + "step": 11025 + }, + { + "epoch": 2.8912188728702493, + "grad_norm": 0.790738046169281, + "learning_rate": 4.474875691451448e-05, + "loss": 0.4028, + "num_input_tokens_seen": 2957528, + "step": 11030 + }, + { + "epoch": 2.892529488859764, + "grad_norm": 0.8393666744232178, + "learning_rate": 4.474174183942186e-05, + "loss": 0.3844, + "num_input_tokens_seen": 2958520, + "step": 11035 + }, + { + "epoch": 2.893840104849279, + "grad_norm": 0.6947301626205444, + "learning_rate": 4.473472263241678e-05, + "loss": 0.6704, + "num_input_tokens_seen": 2959752, + "step": 11040 + }, + { + "epoch": 2.895150720838794, + "grad_norm": 0.9968274235725403, + "learning_rate": 4.472769929496835e-05, + "loss": 0.424, + "num_input_tokens_seen": 2961352, + "step": 11045 + }, + { + "epoch": 2.896461336828309, + "grad_norm": 1.080013394355774, + "learning_rate": 4.472067182854654e-05, + "loss": 0.5798, + "num_input_tokens_seen": 2962568, + "step": 11050 + }, + { + "epoch": 2.8977719528178243, + "grad_norm": 2.097052812576294, + "learning_rate": 4.4713640234622205e-05, + "loss": 0.4802, + "num_input_tokens_seen": 2963816, + "step": 11055 + }, + { + "epoch": 2.8990825688073394, + "grad_norm": 0.35492876172065735, + "learning_rate": 4.4706604514667025e-05, + "loss": 0.4347, + "num_input_tokens_seen": 2965096, + "step": 11060 + }, + { + "epoch": 2.9003931847968545, + "grad_norm": 0.5429349541664124, + "learning_rate": 4.469956467015357e-05, + "loss": 0.603, + "num_input_tokens_seen": 2966744, + "step": 11065 + }, + { + "epoch": 2.9017038007863696, + "grad_norm": 0.9303331971168518, + "learning_rate": 4.469252070255527e-05, + "loss": 0.4973, + "num_input_tokens_seen": 2967768, + "step": 11070 + }, + { + "epoch": 2.9030144167758847, + "grad_norm": 0.46170854568481445, + "learning_rate": 4.4685472613346404e-05, + "loss": 0.5217, + "num_input_tokens_seen": 2969240, + "step": 11075 + }, + { + "epoch": 2.9043250327653998, + "grad_norm": 0.8658671975135803, + "learning_rate": 4.467842040400214e-05, + "loss": 0.4945, + "num_input_tokens_seen": 2970872, + "step": 11080 + }, + { + "epoch": 2.905635648754915, + "grad_norm": 0.26633527874946594, + "learning_rate": 4.467136407599849e-05, + "loss": 0.4036, + "num_input_tokens_seen": 2972296, + "step": 11085 + }, + { + "epoch": 2.90694626474443, + "grad_norm": 0.37450292706489563, + "learning_rate": 4.466430363081233e-05, + "loss": 0.3666, + "num_input_tokens_seen": 2974408, + "step": 11090 + }, + { + "epoch": 2.908256880733945, + "grad_norm": 0.735551118850708, + "learning_rate": 4.465723906992139e-05, + "loss": 0.5346, + "num_input_tokens_seen": 2975448, + "step": 11095 + }, + { + "epoch": 2.90956749672346, + "grad_norm": 1.6154040098190308, + "learning_rate": 4.465017039480428e-05, + "loss": 0.5189, + "num_input_tokens_seen": 2976648, + "step": 11100 + }, + { + "epoch": 2.9108781127129753, + "grad_norm": 0.42729780077934265, + "learning_rate": 4.464309760694047e-05, + "loss": 0.449, + "num_input_tokens_seen": 2978216, + "step": 11105 + }, + { + "epoch": 2.91218872870249, + "grad_norm": 0.7322933673858643, + "learning_rate": 4.4636020707810254e-05, + "loss": 0.4235, + "num_input_tokens_seen": 2979368, + "step": 11110 + }, + { + "epoch": 2.9134993446920054, + "grad_norm": 0.571746826171875, + "learning_rate": 4.462893969889484e-05, + "loss": 0.3697, + "num_input_tokens_seen": 2980760, + "step": 11115 + }, + { + "epoch": 2.91480996068152, + "grad_norm": 1.018829584121704, + "learning_rate": 4.462185458167626e-05, + "loss": 0.4055, + "num_input_tokens_seen": 2982280, + "step": 11120 + }, + { + "epoch": 2.9161205766710356, + "grad_norm": 0.8165537118911743, + "learning_rate": 4.4614765357637425e-05, + "loss": 0.3858, + "num_input_tokens_seen": 2983624, + "step": 11125 + }, + { + "epoch": 2.9174311926605503, + "grad_norm": 0.5480830669403076, + "learning_rate": 4.4607672028262094e-05, + "loss": 0.4453, + "num_input_tokens_seen": 2984936, + "step": 11130 + }, + { + "epoch": 2.9187418086500654, + "grad_norm": 1.315227746963501, + "learning_rate": 4.460057459503488e-05, + "loss": 0.5424, + "num_input_tokens_seen": 2985864, + "step": 11135 + }, + { + "epoch": 2.9200524246395805, + "grad_norm": 0.5405170917510986, + "learning_rate": 4.4593473059441274e-05, + "loss": 0.6238, + "num_input_tokens_seen": 2987336, + "step": 11140 + }, + { + "epoch": 2.9213630406290956, + "grad_norm": 1.305277943611145, + "learning_rate": 4.458636742296761e-05, + "loss": 0.3971, + "num_input_tokens_seen": 2988712, + "step": 11145 + }, + { + "epoch": 2.9226736566186107, + "grad_norm": 1.8767995834350586, + "learning_rate": 4.4579257687101097e-05, + "loss": 0.4692, + "num_input_tokens_seen": 2990184, + "step": 11150 + }, + { + "epoch": 2.9239842726081258, + "grad_norm": 0.33173686265945435, + "learning_rate": 4.457214385332977e-05, + "loss": 0.4076, + "num_input_tokens_seen": 2991720, + "step": 11155 + }, + { + "epoch": 2.925294888597641, + "grad_norm": 0.8372304439544678, + "learning_rate": 4.456502592314256e-05, + "loss": 0.4066, + "num_input_tokens_seen": 2993352, + "step": 11160 + }, + { + "epoch": 2.926605504587156, + "grad_norm": 0.4848769009113312, + "learning_rate": 4.455790389802922e-05, + "loss": 0.4218, + "num_input_tokens_seen": 2995048, + "step": 11165 + }, + { + "epoch": 2.927916120576671, + "grad_norm": 0.6211804747581482, + "learning_rate": 4.4550777779480393e-05, + "loss": 0.3076, + "num_input_tokens_seen": 2996296, + "step": 11170 + }, + { + "epoch": 2.929226736566186, + "grad_norm": 0.5035138726234436, + "learning_rate": 4.454364756898756e-05, + "loss": 0.37, + "num_input_tokens_seen": 2997240, + "step": 11175 + }, + { + "epoch": 2.9305373525557012, + "grad_norm": 0.5565610527992249, + "learning_rate": 4.453651326804305e-05, + "loss": 0.4256, + "num_input_tokens_seen": 2998328, + "step": 11180 + }, + { + "epoch": 2.9318479685452163, + "grad_norm": 0.9306291937828064, + "learning_rate": 4.452937487814007e-05, + "loss": 0.3606, + "num_input_tokens_seen": 2999656, + "step": 11185 + }, + { + "epoch": 2.9331585845347314, + "grad_norm": 0.848270058631897, + "learning_rate": 4.452223240077269e-05, + "loss": 0.2999, + "num_input_tokens_seen": 3000968, + "step": 11190 + }, + { + "epoch": 2.9344692005242465, + "grad_norm": 0.863657534122467, + "learning_rate": 4.451508583743578e-05, + "loss": 0.5905, + "num_input_tokens_seen": 3002696, + "step": 11195 + }, + { + "epoch": 2.9357798165137616, + "grad_norm": 0.7274119257926941, + "learning_rate": 4.450793518962514e-05, + "loss": 0.6629, + "num_input_tokens_seen": 3003832, + "step": 11200 + }, + { + "epoch": 2.9370904325032763, + "grad_norm": 1.2486540079116821, + "learning_rate": 4.450078045883736e-05, + "loss": 0.4392, + "num_input_tokens_seen": 3004984, + "step": 11205 + }, + { + "epoch": 2.938401048492792, + "grad_norm": 1.4802477359771729, + "learning_rate": 4.449362164656993e-05, + "loss": 0.5339, + "num_input_tokens_seen": 3006040, + "step": 11210 + }, + { + "epoch": 2.9397116644823065, + "grad_norm": 0.776604175567627, + "learning_rate": 4.448645875432117e-05, + "loss": 0.4557, + "num_input_tokens_seen": 3007448, + "step": 11215 + }, + { + "epoch": 2.941022280471822, + "grad_norm": 0.41546693444252014, + "learning_rate": 4.447929178359026e-05, + "loss": 0.4614, + "num_input_tokens_seen": 3008776, + "step": 11220 + }, + { + "epoch": 2.9423328964613367, + "grad_norm": 0.6959542036056519, + "learning_rate": 4.4472120735877246e-05, + "loss": 0.4676, + "num_input_tokens_seen": 3009864, + "step": 11225 + }, + { + "epoch": 2.9436435124508518, + "grad_norm": 0.3523676097393036, + "learning_rate": 4.4464945612683005e-05, + "loss": 0.5615, + "num_input_tokens_seen": 3011176, + "step": 11230 + }, + { + "epoch": 2.944954128440367, + "grad_norm": 0.5941627621650696, + "learning_rate": 4.4457766415509275e-05, + "loss": 0.3726, + "num_input_tokens_seen": 3012664, + "step": 11235 + }, + { + "epoch": 2.946264744429882, + "grad_norm": 0.855116069316864, + "learning_rate": 4.445058314585866e-05, + "loss": 0.4599, + "num_input_tokens_seen": 3013704, + "step": 11240 + }, + { + "epoch": 2.947575360419397, + "grad_norm": 0.7862695455551147, + "learning_rate": 4.44433958052346e-05, + "loss": 0.4777, + "num_input_tokens_seen": 3015224, + "step": 11245 + }, + { + "epoch": 2.948885976408912, + "grad_norm": 0.7578448057174683, + "learning_rate": 4.443620439514138e-05, + "loss": 0.5097, + "num_input_tokens_seen": 3016776, + "step": 11250 + }, + { + "epoch": 2.9501965923984272, + "grad_norm": 0.41413435339927673, + "learning_rate": 4.442900891708418e-05, + "loss": 0.5463, + "num_input_tokens_seen": 3017960, + "step": 11255 + }, + { + "epoch": 2.9515072083879423, + "grad_norm": 0.9105470180511475, + "learning_rate": 4.442180937256898e-05, + "loss": 0.3822, + "num_input_tokens_seen": 3019176, + "step": 11260 + }, + { + "epoch": 2.9528178243774574, + "grad_norm": 0.8696861863136292, + "learning_rate": 4.4414605763102635e-05, + "loss": 0.4823, + "num_input_tokens_seen": 3020392, + "step": 11265 + }, + { + "epoch": 2.9541284403669725, + "grad_norm": 0.6005648970603943, + "learning_rate": 4.4407398090192846e-05, + "loss": 0.3543, + "num_input_tokens_seen": 3021496, + "step": 11270 + }, + { + "epoch": 2.9554390563564876, + "grad_norm": 1.3051356077194214, + "learning_rate": 4.440018635534818e-05, + "loss": 0.5694, + "num_input_tokens_seen": 3022536, + "step": 11275 + }, + { + "epoch": 2.9567496723460027, + "grad_norm": 2.4365592002868652, + "learning_rate": 4.439297056007802e-05, + "loss": 0.6251, + "num_input_tokens_seen": 3024024, + "step": 11280 + }, + { + "epoch": 2.958060288335518, + "grad_norm": 0.9987600445747375, + "learning_rate": 4.4385750705892634e-05, + "loss": 0.4259, + "num_input_tokens_seen": 3025256, + "step": 11285 + }, + { + "epoch": 2.959370904325033, + "grad_norm": 1.0821572542190552, + "learning_rate": 4.437852679430313e-05, + "loss": 0.3884, + "num_input_tokens_seen": 3026792, + "step": 11290 + }, + { + "epoch": 2.960681520314548, + "grad_norm": 0.7632647156715393, + "learning_rate": 4.437129882682144e-05, + "loss": 0.4688, + "num_input_tokens_seen": 3028424, + "step": 11295 + }, + { + "epoch": 2.9619921363040627, + "grad_norm": 0.791169285774231, + "learning_rate": 4.436406680496038e-05, + "loss": 0.4044, + "num_input_tokens_seen": 3029416, + "step": 11300 + }, + { + "epoch": 2.963302752293578, + "grad_norm": 0.6125750541687012, + "learning_rate": 4.43568307302336e-05, + "loss": 0.4301, + "num_input_tokens_seen": 3030776, + "step": 11305 + }, + { + "epoch": 2.964613368283093, + "grad_norm": 1.383603572845459, + "learning_rate": 4.4349590604155586e-05, + "loss": 0.4353, + "num_input_tokens_seen": 3031896, + "step": 11310 + }, + { + "epoch": 2.9659239842726084, + "grad_norm": 1.690375804901123, + "learning_rate": 4.43423464282417e-05, + "loss": 0.59, + "num_input_tokens_seen": 3032952, + "step": 11315 + }, + { + "epoch": 2.967234600262123, + "grad_norm": 0.8841455578804016, + "learning_rate": 4.4335098204008126e-05, + "loss": 0.5264, + "num_input_tokens_seen": 3034760, + "step": 11320 + }, + { + "epoch": 2.968545216251638, + "grad_norm": 0.8409136533737183, + "learning_rate": 4.43278459329719e-05, + "loss": 0.572, + "num_input_tokens_seen": 3035976, + "step": 11325 + }, + { + "epoch": 2.9698558322411532, + "grad_norm": 0.4639868140220642, + "learning_rate": 4.432058961665092e-05, + "loss": 0.5973, + "num_input_tokens_seen": 3037304, + "step": 11330 + }, + { + "epoch": 2.9711664482306683, + "grad_norm": 0.7674406170845032, + "learning_rate": 4.4313329256563916e-05, + "loss": 0.3219, + "num_input_tokens_seen": 3038664, + "step": 11335 + }, + { + "epoch": 2.9724770642201834, + "grad_norm": 0.5720927715301514, + "learning_rate": 4.430606485423048e-05, + "loss": 0.3878, + "num_input_tokens_seen": 3040024, + "step": 11340 + }, + { + "epoch": 2.9737876802096985, + "grad_norm": 0.4780954420566559, + "learning_rate": 4.4298796411171015e-05, + "loss": 0.5517, + "num_input_tokens_seen": 3041320, + "step": 11345 + }, + { + "epoch": 2.9750982961992136, + "grad_norm": 0.5104770064353943, + "learning_rate": 4.429152392890681e-05, + "loss": 0.4018, + "num_input_tokens_seen": 3042408, + "step": 11350 + }, + { + "epoch": 2.9764089121887287, + "grad_norm": 0.7084235548973083, + "learning_rate": 4.428424740895998e-05, + "loss": 0.4687, + "num_input_tokens_seen": 3043640, + "step": 11355 + }, + { + "epoch": 2.977719528178244, + "grad_norm": 0.7815394401550293, + "learning_rate": 4.427696685285349e-05, + "loss": 0.4836, + "num_input_tokens_seen": 3045176, + "step": 11360 + }, + { + "epoch": 2.979030144167759, + "grad_norm": 0.480859637260437, + "learning_rate": 4.4269682262111145e-05, + "loss": 0.3748, + "num_input_tokens_seen": 3046504, + "step": 11365 + }, + { + "epoch": 2.980340760157274, + "grad_norm": 4.285477638244629, + "learning_rate": 4.42623936382576e-05, + "loss": 0.3986, + "num_input_tokens_seen": 3047752, + "step": 11370 + }, + { + "epoch": 2.981651376146789, + "grad_norm": 52.59010314941406, + "learning_rate": 4.425510098281835e-05, + "loss": 0.6448, + "num_input_tokens_seen": 3049368, + "step": 11375 + }, + { + "epoch": 2.982961992136304, + "grad_norm": 0.7550553679466248, + "learning_rate": 4.424780429731973e-05, + "loss": 0.4639, + "num_input_tokens_seen": 3050712, + "step": 11380 + }, + { + "epoch": 2.9842726081258193, + "grad_norm": 0.8583986163139343, + "learning_rate": 4.424050358328893e-05, + "loss": 0.5218, + "num_input_tokens_seen": 3052488, + "step": 11385 + }, + { + "epoch": 2.9855832241153344, + "grad_norm": 1.4429796934127808, + "learning_rate": 4.423319884225398e-05, + "loss": 0.4873, + "num_input_tokens_seen": 3053736, + "step": 11390 + }, + { + "epoch": 2.986893840104849, + "grad_norm": 1.2058653831481934, + "learning_rate": 4.422589007574374e-05, + "loss": 0.331, + "num_input_tokens_seen": 3054856, + "step": 11395 + }, + { + "epoch": 2.9882044560943646, + "grad_norm": 1.0005481243133545, + "learning_rate": 4.421857728528792e-05, + "loss": 0.3954, + "num_input_tokens_seen": 3055848, + "step": 11400 + }, + { + "epoch": 2.9895150720838792, + "grad_norm": 0.48381397128105164, + "learning_rate": 4.421126047241709e-05, + "loss": 0.4025, + "num_input_tokens_seen": 3057112, + "step": 11405 + }, + { + "epoch": 2.9908256880733948, + "grad_norm": 0.614581286907196, + "learning_rate": 4.420393963866263e-05, + "loss": 0.5234, + "num_input_tokens_seen": 3058312, + "step": 11410 + }, + { + "epoch": 2.9921363040629094, + "grad_norm": 0.7452767491340637, + "learning_rate": 4.4196614785556795e-05, + "loss": 0.4033, + "num_input_tokens_seen": 3060008, + "step": 11415 + }, + { + "epoch": 2.9934469200524245, + "grad_norm": 3.896138906478882, + "learning_rate": 4.418928591463265e-05, + "loss": 0.4826, + "num_input_tokens_seen": 3061016, + "step": 11420 + }, + { + "epoch": 2.9947575360419396, + "grad_norm": 0.42438310384750366, + "learning_rate": 4.418195302742412e-05, + "loss": 0.2875, + "num_input_tokens_seen": 3062632, + "step": 11425 + }, + { + "epoch": 2.9960681520314547, + "grad_norm": 0.5392744541168213, + "learning_rate": 4.417461612546596e-05, + "loss": 0.4884, + "num_input_tokens_seen": 3064008, + "step": 11430 + }, + { + "epoch": 2.99737876802097, + "grad_norm": 1.93932044506073, + "learning_rate": 4.416727521029379e-05, + "loss": 0.4887, + "num_input_tokens_seen": 3065256, + "step": 11435 + }, + { + "epoch": 2.998689384010485, + "grad_norm": 0.8925154805183411, + "learning_rate": 4.415993028344403e-05, + "loss": 0.393, + "num_input_tokens_seen": 3066456, + "step": 11440 + }, + { + "epoch": 3.0, + "grad_norm": 3.5052998065948486, + "learning_rate": 4.415258134645396e-05, + "loss": 0.3892, + "num_input_tokens_seen": 3067768, + "step": 11445 + }, + { + "epoch": 3.000786369593709, + "eval_loss": 0.4982500374317169, + "eval_runtime": 16.8552, + "eval_samples_per_second": 50.311, + "eval_steps_per_second": 25.155, + "num_input_tokens_seen": 3068424, + "step": 11448 + }, + { + "epoch": 3.001310615989515, + "grad_norm": 0.7516046762466431, + "learning_rate": 4.414522840086172e-05, + "loss": 0.5965, + "num_input_tokens_seen": 3068920, + "step": 11450 + }, + { + "epoch": 3.00262123197903, + "grad_norm": 0.6731840372085571, + "learning_rate": 4.413787144820625e-05, + "loss": 0.4483, + "num_input_tokens_seen": 3070328, + "step": 11455 + }, + { + "epoch": 3.0039318479685453, + "grad_norm": 2.0179810523986816, + "learning_rate": 4.413051049002735e-05, + "loss": 0.3816, + "num_input_tokens_seen": 3071368, + "step": 11460 + }, + { + "epoch": 3.0052424639580604, + "grad_norm": 0.9936145544052124, + "learning_rate": 4.412314552786566e-05, + "loss": 0.4642, + "num_input_tokens_seen": 3072840, + "step": 11465 + }, + { + "epoch": 3.0065530799475755, + "grad_norm": 0.6801871657371521, + "learning_rate": 4.4115776563262656e-05, + "loss": 0.3447, + "num_input_tokens_seen": 3074440, + "step": 11470 + }, + { + "epoch": 3.0078636959370906, + "grad_norm": 1.6887075901031494, + "learning_rate": 4.410840359776065e-05, + "loss": 0.3527, + "num_input_tokens_seen": 3075592, + "step": 11475 + }, + { + "epoch": 3.0091743119266057, + "grad_norm": 1.3335708379745483, + "learning_rate": 4.410102663290278e-05, + "loss": 0.4124, + "num_input_tokens_seen": 3076616, + "step": 11480 + }, + { + "epoch": 3.0104849279161208, + "grad_norm": 0.4669129252433777, + "learning_rate": 4.4093645670233045e-05, + "loss": 0.3767, + "num_input_tokens_seen": 3078088, + "step": 11485 + }, + { + "epoch": 3.011795543905636, + "grad_norm": 1.3585636615753174, + "learning_rate": 4.4086260711296265e-05, + "loss": 0.4629, + "num_input_tokens_seen": 3079064, + "step": 11490 + }, + { + "epoch": 3.0131061598951505, + "grad_norm": 0.9167100787162781, + "learning_rate": 4.407887175763809e-05, + "loss": 0.3462, + "num_input_tokens_seen": 3080184, + "step": 11495 + }, + { + "epoch": 3.0144167758846656, + "grad_norm": 0.389627069234848, + "learning_rate": 4.407147881080503e-05, + "loss": 0.3782, + "num_input_tokens_seen": 3082008, + "step": 11500 + }, + { + "epoch": 3.0157273918741807, + "grad_norm": 0.7433416247367859, + "learning_rate": 4.40640818723444e-05, + "loss": 0.4333, + "num_input_tokens_seen": 3083352, + "step": 11505 + }, + { + "epoch": 3.017038007863696, + "grad_norm": 1.3437498807907104, + "learning_rate": 4.4056680943804374e-05, + "loss": 0.3723, + "num_input_tokens_seen": 3084248, + "step": 11510 + }, + { + "epoch": 3.018348623853211, + "grad_norm": 0.6837265491485596, + "learning_rate": 4.404927602673396e-05, + "loss": 0.4296, + "num_input_tokens_seen": 3085864, + "step": 11515 + }, + { + "epoch": 3.019659239842726, + "grad_norm": 1.1469885110855103, + "learning_rate": 4.404186712268299e-05, + "loss": 0.4765, + "num_input_tokens_seen": 3087112, + "step": 11520 + }, + { + "epoch": 3.020969855832241, + "grad_norm": 0.7509775757789612, + "learning_rate": 4.4034454233202124e-05, + "loss": 0.2866, + "num_input_tokens_seen": 3088632, + "step": 11525 + }, + { + "epoch": 3.022280471821756, + "grad_norm": 0.5605435371398926, + "learning_rate": 4.4027037359842884e-05, + "loss": 0.393, + "num_input_tokens_seen": 3089560, + "step": 11530 + }, + { + "epoch": 3.0235910878112713, + "grad_norm": 0.7636898756027222, + "learning_rate": 4.401961650415759e-05, + "loss": 0.3265, + "num_input_tokens_seen": 3090600, + "step": 11535 + }, + { + "epoch": 3.0249017038007864, + "grad_norm": 0.7916657328605652, + "learning_rate": 4.401219166769943e-05, + "loss": 0.3968, + "num_input_tokens_seen": 3091704, + "step": 11540 + }, + { + "epoch": 3.0262123197903015, + "grad_norm": 4.296518325805664, + "learning_rate": 4.400476285202241e-05, + "loss": 0.4352, + "num_input_tokens_seen": 3092744, + "step": 11545 + }, + { + "epoch": 3.0275229357798166, + "grad_norm": 0.6140381693840027, + "learning_rate": 4.399733005868136e-05, + "loss": 0.506, + "num_input_tokens_seen": 3094408, + "step": 11550 + }, + { + "epoch": 3.0288335517693317, + "grad_norm": 0.7647310495376587, + "learning_rate": 4.3989893289231954e-05, + "loss": 0.3539, + "num_input_tokens_seen": 3095800, + "step": 11555 + }, + { + "epoch": 3.0301441677588468, + "grad_norm": 0.7629491090774536, + "learning_rate": 4.398245254523069e-05, + "loss": 0.5597, + "num_input_tokens_seen": 3096888, + "step": 11560 + }, + { + "epoch": 3.031454783748362, + "grad_norm": 0.8275530934333801, + "learning_rate": 4.397500782823492e-05, + "loss": 0.6472, + "num_input_tokens_seen": 3098856, + "step": 11565 + }, + { + "epoch": 3.032765399737877, + "grad_norm": 0.8514381647109985, + "learning_rate": 4.3967559139802785e-05, + "loss": 0.3778, + "num_input_tokens_seen": 3100184, + "step": 11570 + }, + { + "epoch": 3.034076015727392, + "grad_norm": 0.3639926612377167, + "learning_rate": 4.39601064814933e-05, + "loss": 0.4073, + "num_input_tokens_seen": 3101688, + "step": 11575 + }, + { + "epoch": 3.035386631716907, + "grad_norm": 0.4975484311580658, + "learning_rate": 4.395264985486629e-05, + "loss": 0.3867, + "num_input_tokens_seen": 3103048, + "step": 11580 + }, + { + "epoch": 3.036697247706422, + "grad_norm": 0.5522270202636719, + "learning_rate": 4.394518926148241e-05, + "loss": 0.3238, + "num_input_tokens_seen": 3104376, + "step": 11585 + }, + { + "epoch": 3.038007863695937, + "grad_norm": 0.6053511500358582, + "learning_rate": 4.393772470290315e-05, + "loss": 0.4289, + "num_input_tokens_seen": 3105720, + "step": 11590 + }, + { + "epoch": 3.039318479685452, + "grad_norm": 0.8523383736610413, + "learning_rate": 4.393025618069084e-05, + "loss": 0.4933, + "num_input_tokens_seen": 3107080, + "step": 11595 + }, + { + "epoch": 3.040629095674967, + "grad_norm": 1.1332203149795532, + "learning_rate": 4.3922783696408615e-05, + "loss": 0.3588, + "num_input_tokens_seen": 3108648, + "step": 11600 + }, + { + "epoch": 3.041939711664482, + "grad_norm": 0.4790290296077728, + "learning_rate": 4.3915307251620456e-05, + "loss": 0.4388, + "num_input_tokens_seen": 3110040, + "step": 11605 + }, + { + "epoch": 3.0432503276539973, + "grad_norm": 1.5411278009414673, + "learning_rate": 4.390782684789118e-05, + "loss": 0.4702, + "num_input_tokens_seen": 3111672, + "step": 11610 + }, + { + "epoch": 3.0445609436435124, + "grad_norm": 0.918713390827179, + "learning_rate": 4.3900342486786405e-05, + "loss": 0.4637, + "num_input_tokens_seen": 3112936, + "step": 11615 + }, + { + "epoch": 3.0458715596330275, + "grad_norm": 1.7030304670333862, + "learning_rate": 4.389285416987261e-05, + "loss": 0.4431, + "num_input_tokens_seen": 3113928, + "step": 11620 + }, + { + "epoch": 3.0471821756225426, + "grad_norm": 0.9203720092773438, + "learning_rate": 4.388536189871708e-05, + "loss": 0.3941, + "num_input_tokens_seen": 3115384, + "step": 11625 + }, + { + "epoch": 3.0484927916120577, + "grad_norm": 0.6991195678710938, + "learning_rate": 4.387786567488794e-05, + "loss": 0.9114, + "num_input_tokens_seen": 3116616, + "step": 11630 + }, + { + "epoch": 3.0498034076015728, + "grad_norm": 1.0890594720840454, + "learning_rate": 4.387036549995413e-05, + "loss": 0.3731, + "num_input_tokens_seen": 3117768, + "step": 11635 + }, + { + "epoch": 3.051114023591088, + "grad_norm": 1.1694154739379883, + "learning_rate": 4.386286137548543e-05, + "loss": 0.3937, + "num_input_tokens_seen": 3119064, + "step": 11640 + }, + { + "epoch": 3.052424639580603, + "grad_norm": 0.8926483988761902, + "learning_rate": 4.3855353303052436e-05, + "loss": 0.5141, + "num_input_tokens_seen": 3120520, + "step": 11645 + }, + { + "epoch": 3.053735255570118, + "grad_norm": 1.1834152936935425, + "learning_rate": 4.384784128422657e-05, + "loss": 0.4238, + "num_input_tokens_seen": 3121416, + "step": 11650 + }, + { + "epoch": 3.055045871559633, + "grad_norm": 0.8706555962562561, + "learning_rate": 4.3840325320580086e-05, + "loss": 0.5727, + "num_input_tokens_seen": 3122744, + "step": 11655 + }, + { + "epoch": 3.0563564875491482, + "grad_norm": 0.616183340549469, + "learning_rate": 4.383280541368608e-05, + "loss": 0.3003, + "num_input_tokens_seen": 3125080, + "step": 11660 + }, + { + "epoch": 3.0576671035386633, + "grad_norm": 0.47159865498542786, + "learning_rate": 4.382528156511843e-05, + "loss": 0.3629, + "num_input_tokens_seen": 3126104, + "step": 11665 + }, + { + "epoch": 3.0589777195281784, + "grad_norm": 0.9636691212654114, + "learning_rate": 4.381775377645188e-05, + "loss": 0.4208, + "num_input_tokens_seen": 3127304, + "step": 11670 + }, + { + "epoch": 3.0602883355176935, + "grad_norm": 1.05583918094635, + "learning_rate": 4.381022204926198e-05, + "loss": 0.3621, + "num_input_tokens_seen": 3128712, + "step": 11675 + }, + { + "epoch": 3.061598951507208, + "grad_norm": 0.6097434759140015, + "learning_rate": 4.38026863851251e-05, + "loss": 0.6157, + "num_input_tokens_seen": 3129720, + "step": 11680 + }, + { + "epoch": 3.0629095674967233, + "grad_norm": 0.4235377907752991, + "learning_rate": 4.379514678561846e-05, + "loss": 0.4427, + "num_input_tokens_seen": 3131224, + "step": 11685 + }, + { + "epoch": 3.0642201834862384, + "grad_norm": 0.43919429183006287, + "learning_rate": 4.3787603252320065e-05, + "loss": 0.2769, + "num_input_tokens_seen": 3132888, + "step": 11690 + }, + { + "epoch": 3.0655307994757535, + "grad_norm": 0.7082739472389221, + "learning_rate": 4.378005578680877e-05, + "loss": 0.388, + "num_input_tokens_seen": 3134472, + "step": 11695 + }, + { + "epoch": 3.0668414154652686, + "grad_norm": 2.1977593898773193, + "learning_rate": 4.377250439066426e-05, + "loss": 0.5551, + "num_input_tokens_seen": 3135416, + "step": 11700 + }, + { + "epoch": 3.0681520314547837, + "grad_norm": 0.8497104048728943, + "learning_rate": 4.3764949065467006e-05, + "loss": 0.769, + "num_input_tokens_seen": 3136616, + "step": 11705 + }, + { + "epoch": 3.0694626474442988, + "grad_norm": 1.4423209428787231, + "learning_rate": 4.375738981279834e-05, + "loss": 0.4483, + "num_input_tokens_seen": 3137720, + "step": 11710 + }, + { + "epoch": 3.070773263433814, + "grad_norm": 0.6646153330802917, + "learning_rate": 4.37498266342404e-05, + "loss": 0.4704, + "num_input_tokens_seen": 3138680, + "step": 11715 + }, + { + "epoch": 3.072083879423329, + "grad_norm": 0.5779693722724915, + "learning_rate": 4.3742259531376135e-05, + "loss": 0.252, + "num_input_tokens_seen": 3139672, + "step": 11720 + }, + { + "epoch": 3.073394495412844, + "grad_norm": 0.9206782579421997, + "learning_rate": 4.373468850578933e-05, + "loss": 0.249, + "num_input_tokens_seen": 3140824, + "step": 11725 + }, + { + "epoch": 3.074705111402359, + "grad_norm": 2.180929183959961, + "learning_rate": 4.37271135590646e-05, + "loss": 0.4372, + "num_input_tokens_seen": 3142040, + "step": 11730 + }, + { + "epoch": 3.0760157273918742, + "grad_norm": 0.3608691394329071, + "learning_rate": 4.371953469278737e-05, + "loss": 0.1898, + "num_input_tokens_seen": 3143080, + "step": 11735 + }, + { + "epoch": 3.0773263433813893, + "grad_norm": 0.724331259727478, + "learning_rate": 4.371195190854386e-05, + "loss": 0.3967, + "num_input_tokens_seen": 3144504, + "step": 11740 + }, + { + "epoch": 3.0786369593709044, + "grad_norm": 0.725860595703125, + "learning_rate": 4.370436520792115e-05, + "loss": 0.4637, + "num_input_tokens_seen": 3145816, + "step": 11745 + }, + { + "epoch": 3.0799475753604195, + "grad_norm": 3.8819689750671387, + "learning_rate": 4.369677459250712e-05, + "loss": 0.5318, + "num_input_tokens_seen": 3147528, + "step": 11750 + }, + { + "epoch": 3.0812581913499346, + "grad_norm": 1.476914644241333, + "learning_rate": 4.3689180063890476e-05, + "loss": 0.526, + "num_input_tokens_seen": 3148648, + "step": 11755 + }, + { + "epoch": 3.0825688073394497, + "grad_norm": 0.8274660706520081, + "learning_rate": 4.368158162366073e-05, + "loss": 0.364, + "num_input_tokens_seen": 3149976, + "step": 11760 + }, + { + "epoch": 3.083879423328965, + "grad_norm": 2.547461986541748, + "learning_rate": 4.367397927340824e-05, + "loss": 0.5678, + "num_input_tokens_seen": 3151624, + "step": 11765 + }, + { + "epoch": 3.08519003931848, + "grad_norm": 1.5462589263916016, + "learning_rate": 4.366637301472415e-05, + "loss": 0.4596, + "num_input_tokens_seen": 3153160, + "step": 11770 + }, + { + "epoch": 3.0865006553079946, + "grad_norm": 0.7571626901626587, + "learning_rate": 4.365876284920044e-05, + "loss": 0.4235, + "num_input_tokens_seen": 3154600, + "step": 11775 + }, + { + "epoch": 3.0878112712975097, + "grad_norm": 0.8172830939292908, + "learning_rate": 4.365114877842991e-05, + "loss": 0.298, + "num_input_tokens_seen": 3155912, + "step": 11780 + }, + { + "epoch": 3.0891218872870247, + "grad_norm": 1.5201280117034912, + "learning_rate": 4.364353080400616e-05, + "loss": 0.473, + "num_input_tokens_seen": 3156936, + "step": 11785 + }, + { + "epoch": 3.09043250327654, + "grad_norm": 0.9551549553871155, + "learning_rate": 4.363590892752363e-05, + "loss": 0.413, + "num_input_tokens_seen": 3158216, + "step": 11790 + }, + { + "epoch": 3.091743119266055, + "grad_norm": 0.5680015683174133, + "learning_rate": 4.362828315057756e-05, + "loss": 0.4592, + "num_input_tokens_seen": 3159544, + "step": 11795 + }, + { + "epoch": 3.09305373525557, + "grad_norm": 0.7975147366523743, + "learning_rate": 4.362065347476401e-05, + "loss": 0.3698, + "num_input_tokens_seen": 3161384, + "step": 11800 + }, + { + "epoch": 3.094364351245085, + "grad_norm": 0.5364548563957214, + "learning_rate": 4.361301990167988e-05, + "loss": 0.2577, + "num_input_tokens_seen": 3162360, + "step": 11805 + }, + { + "epoch": 3.0956749672346002, + "grad_norm": 1.1497511863708496, + "learning_rate": 4.360538243292282e-05, + "loss": 0.4642, + "num_input_tokens_seen": 3163352, + "step": 11810 + }, + { + "epoch": 3.0969855832241153, + "grad_norm": 0.5623584389686584, + "learning_rate": 4.359774107009138e-05, + "loss": 0.3416, + "num_input_tokens_seen": 3164648, + "step": 11815 + }, + { + "epoch": 3.0982961992136304, + "grad_norm": 1.2053474187850952, + "learning_rate": 4.3590095814784866e-05, + "loss": 0.3921, + "num_input_tokens_seen": 3165752, + "step": 11820 + }, + { + "epoch": 3.0996068152031455, + "grad_norm": 1.157063603401184, + "learning_rate": 4.3582446668603414e-05, + "loss": 0.2985, + "num_input_tokens_seen": 3167384, + "step": 11825 + }, + { + "epoch": 3.1009174311926606, + "grad_norm": 0.5165517330169678, + "learning_rate": 4.357479363314799e-05, + "loss": 0.3986, + "num_input_tokens_seen": 3169288, + "step": 11830 + }, + { + "epoch": 3.1022280471821757, + "grad_norm": 1.0864025354385376, + "learning_rate": 4.356713671002035e-05, + "loss": 0.3172, + "num_input_tokens_seen": 3170264, + "step": 11835 + }, + { + "epoch": 3.103538663171691, + "grad_norm": 1.092415452003479, + "learning_rate": 4.355947590082308e-05, + "loss": 0.3793, + "num_input_tokens_seen": 3171448, + "step": 11840 + }, + { + "epoch": 3.104849279161206, + "grad_norm": 1.046007752418518, + "learning_rate": 4.355181120715957e-05, + "loss": 0.4418, + "num_input_tokens_seen": 3173016, + "step": 11845 + }, + { + "epoch": 3.106159895150721, + "grad_norm": 0.5127217173576355, + "learning_rate": 4.3544142630634024e-05, + "loss": 0.3108, + "num_input_tokens_seen": 3174872, + "step": 11850 + }, + { + "epoch": 3.107470511140236, + "grad_norm": 0.7208882570266724, + "learning_rate": 4.3536470172851476e-05, + "loss": 0.3533, + "num_input_tokens_seen": 3176840, + "step": 11855 + }, + { + "epoch": 3.108781127129751, + "grad_norm": 0.9273102283477783, + "learning_rate": 4.352879383541775e-05, + "loss": 0.4659, + "num_input_tokens_seen": 3178328, + "step": 11860 + }, + { + "epoch": 3.1100917431192663, + "grad_norm": 0.5803025364875793, + "learning_rate": 4.352111361993948e-05, + "loss": 0.4544, + "num_input_tokens_seen": 3179672, + "step": 11865 + }, + { + "epoch": 3.111402359108781, + "grad_norm": 2.4944026470184326, + "learning_rate": 4.351342952802414e-05, + "loss": 0.4396, + "num_input_tokens_seen": 3181128, + "step": 11870 + }, + { + "epoch": 3.112712975098296, + "grad_norm": 1.8797571659088135, + "learning_rate": 4.3505741561279984e-05, + "loss": 0.2446, + "num_input_tokens_seen": 3182648, + "step": 11875 + }, + { + "epoch": 3.114023591087811, + "grad_norm": 1.131805181503296, + "learning_rate": 4.349804972131609e-05, + "loss": 0.387, + "num_input_tokens_seen": 3184312, + "step": 11880 + }, + { + "epoch": 3.1153342070773262, + "grad_norm": 2.122825860977173, + "learning_rate": 4.349035400974236e-05, + "loss": 0.5017, + "num_input_tokens_seen": 3185448, + "step": 11885 + }, + { + "epoch": 3.1166448230668413, + "grad_norm": 0.818408191204071, + "learning_rate": 4.348265442816949e-05, + "loss": 0.3177, + "num_input_tokens_seen": 3186920, + "step": 11890 + }, + { + "epoch": 3.1179554390563564, + "grad_norm": 0.7421442866325378, + "learning_rate": 4.347495097820897e-05, + "loss": 0.4904, + "num_input_tokens_seen": 3188568, + "step": 11895 + }, + { + "epoch": 3.1192660550458715, + "grad_norm": 0.8635108470916748, + "learning_rate": 4.346724366147314e-05, + "loss": 0.3734, + "num_input_tokens_seen": 3189800, + "step": 11900 + }, + { + "epoch": 3.1205766710353866, + "grad_norm": 1.7708837985992432, + "learning_rate": 4.345953247957512e-05, + "loss": 0.2996, + "num_input_tokens_seen": 3190856, + "step": 11905 + }, + { + "epoch": 3.1218872870249017, + "grad_norm": 0.6710953712463379, + "learning_rate": 4.345181743412885e-05, + "loss": 0.3528, + "num_input_tokens_seen": 3192552, + "step": 11910 + }, + { + "epoch": 3.123197903014417, + "grad_norm": 0.999070405960083, + "learning_rate": 4.344409852674907e-05, + "loss": 0.4727, + "num_input_tokens_seen": 3193672, + "step": 11915 + }, + { + "epoch": 3.124508519003932, + "grad_norm": 0.7589579224586487, + "learning_rate": 4.343637575905133e-05, + "loss": 0.3412, + "num_input_tokens_seen": 3194760, + "step": 11920 + }, + { + "epoch": 3.125819134993447, + "grad_norm": 1.0297437906265259, + "learning_rate": 4.342864913265201e-05, + "loss": 0.2603, + "num_input_tokens_seen": 3195768, + "step": 11925 + }, + { + "epoch": 3.127129750982962, + "grad_norm": 5.361533164978027, + "learning_rate": 4.3420918649168265e-05, + "loss": 0.4667, + "num_input_tokens_seen": 3196936, + "step": 11930 + }, + { + "epoch": 3.128440366972477, + "grad_norm": 2.3005378246307373, + "learning_rate": 4.341318431021808e-05, + "loss": 0.4124, + "num_input_tokens_seen": 3198360, + "step": 11935 + }, + { + "epoch": 3.1297509829619923, + "grad_norm": 0.8609933257102966, + "learning_rate": 4.340544611742023e-05, + "loss": 0.4426, + "num_input_tokens_seen": 3199816, + "step": 11940 + }, + { + "epoch": 3.1310615989515074, + "grad_norm": 0.6249607801437378, + "learning_rate": 4.339770407239431e-05, + "loss": 0.4121, + "num_input_tokens_seen": 3200808, + "step": 11945 + }, + { + "epoch": 3.1323722149410225, + "grad_norm": 0.5043487548828125, + "learning_rate": 4.3389958176760714e-05, + "loss": 0.5386, + "num_input_tokens_seen": 3202984, + "step": 11950 + }, + { + "epoch": 3.133682830930537, + "grad_norm": 0.5435835719108582, + "learning_rate": 4.3382208432140646e-05, + "loss": 0.3013, + "num_input_tokens_seen": 3204296, + "step": 11955 + }, + { + "epoch": 3.134993446920052, + "grad_norm": 1.12422776222229, + "learning_rate": 4.337445484015613e-05, + "loss": 0.371, + "num_input_tokens_seen": 3205704, + "step": 11960 + }, + { + "epoch": 3.1363040629095673, + "grad_norm": 2.8334920406341553, + "learning_rate": 4.3366697402429945e-05, + "loss": 0.4403, + "num_input_tokens_seen": 3207240, + "step": 11965 + }, + { + "epoch": 3.1376146788990824, + "grad_norm": 0.9808211326599121, + "learning_rate": 4.335893612058574e-05, + "loss": 0.389, + "num_input_tokens_seen": 3208264, + "step": 11970 + }, + { + "epoch": 3.1389252948885975, + "grad_norm": 2.8058247566223145, + "learning_rate": 4.335117099624792e-05, + "loss": 0.4544, + "num_input_tokens_seen": 3209656, + "step": 11975 + }, + { + "epoch": 3.1402359108781126, + "grad_norm": 0.8045088052749634, + "learning_rate": 4.3343402031041726e-05, + "loss": 0.3482, + "num_input_tokens_seen": 3210616, + "step": 11980 + }, + { + "epoch": 3.1415465268676277, + "grad_norm": 1.4433207511901855, + "learning_rate": 4.333562922659318e-05, + "loss": 0.4139, + "num_input_tokens_seen": 3211640, + "step": 11985 + }, + { + "epoch": 3.142857142857143, + "grad_norm": 0.7814003229141235, + "learning_rate": 4.332785258452911e-05, + "loss": 0.3847, + "num_input_tokens_seen": 3213096, + "step": 11990 + }, + { + "epoch": 3.144167758846658, + "grad_norm": 1.1052695512771606, + "learning_rate": 4.332007210647717e-05, + "loss": 0.3989, + "num_input_tokens_seen": 3214072, + "step": 11995 + }, + { + "epoch": 3.145478374836173, + "grad_norm": 1.0672801733016968, + "learning_rate": 4.3312287794065795e-05, + "loss": 0.3849, + "num_input_tokens_seen": 3215416, + "step": 12000 + }, + { + "epoch": 3.146788990825688, + "grad_norm": 4.766317367553711, + "learning_rate": 4.330449964892422e-05, + "loss": 0.2404, + "num_input_tokens_seen": 3216680, + "step": 12005 + }, + { + "epoch": 3.148099606815203, + "grad_norm": 1.0426746606826782, + "learning_rate": 4.32967076726825e-05, + "loss": 0.4377, + "num_input_tokens_seen": 3217704, + "step": 12010 + }, + { + "epoch": 3.1494102228047183, + "grad_norm": 1.7925677299499512, + "learning_rate": 4.328891186697148e-05, + "loss": 0.5057, + "num_input_tokens_seen": 3218840, + "step": 12015 + }, + { + "epoch": 3.1507208387942334, + "grad_norm": 1.263033151626587, + "learning_rate": 4.328111223342281e-05, + "loss": 0.2536, + "num_input_tokens_seen": 3220264, + "step": 12020 + }, + { + "epoch": 3.1520314547837485, + "grad_norm": 0.7444867491722107, + "learning_rate": 4.327330877366893e-05, + "loss": 0.8794, + "num_input_tokens_seen": 3221416, + "step": 12025 + }, + { + "epoch": 3.1533420707732636, + "grad_norm": 0.9365726113319397, + "learning_rate": 4.3265501489343094e-05, + "loss": 0.5601, + "num_input_tokens_seen": 3222616, + "step": 12030 + }, + { + "epoch": 3.1546526867627787, + "grad_norm": 0.8694643378257751, + "learning_rate": 4.325769038207936e-05, + "loss": 0.3504, + "num_input_tokens_seen": 3223624, + "step": 12035 + }, + { + "epoch": 3.1559633027522938, + "grad_norm": 0.460722953081131, + "learning_rate": 4.3249875453512586e-05, + "loss": 0.3147, + "num_input_tokens_seen": 3225256, + "step": 12040 + }, + { + "epoch": 3.157273918741809, + "grad_norm": 1.3838565349578857, + "learning_rate": 4.32420567052784e-05, + "loss": 0.4685, + "num_input_tokens_seen": 3226632, + "step": 12045 + }, + { + "epoch": 3.1585845347313235, + "grad_norm": 1.294411540031433, + "learning_rate": 4.323423413901327e-05, + "loss": 0.3407, + "num_input_tokens_seen": 3227768, + "step": 12050 + }, + { + "epoch": 3.1598951507208386, + "grad_norm": 0.5001734495162964, + "learning_rate": 4.322640775635445e-05, + "loss": 0.3802, + "num_input_tokens_seen": 3229128, + "step": 12055 + }, + { + "epoch": 3.1612057667103537, + "grad_norm": 1.3893282413482666, + "learning_rate": 4.321857755893996e-05, + "loss": 0.432, + "num_input_tokens_seen": 3230216, + "step": 12060 + }, + { + "epoch": 3.162516382699869, + "grad_norm": 1.5065678358078003, + "learning_rate": 4.3210743548408684e-05, + "loss": 0.3695, + "num_input_tokens_seen": 3231256, + "step": 12065 + }, + { + "epoch": 3.163826998689384, + "grad_norm": 1.0566908121109009, + "learning_rate": 4.3202905726400236e-05, + "loss": 0.5679, + "num_input_tokens_seen": 3232808, + "step": 12070 + }, + { + "epoch": 3.165137614678899, + "grad_norm": 0.7591590881347656, + "learning_rate": 4.3195064094555085e-05, + "loss": 0.3818, + "num_input_tokens_seen": 3234776, + "step": 12075 + }, + { + "epoch": 3.166448230668414, + "grad_norm": 9.268171310424805, + "learning_rate": 4.318721865451445e-05, + "loss": 0.5757, + "num_input_tokens_seen": 3236088, + "step": 12080 + }, + { + "epoch": 3.167758846657929, + "grad_norm": 0.946866512298584, + "learning_rate": 4.317936940792037e-05, + "loss": 0.4092, + "num_input_tokens_seen": 3237448, + "step": 12085 + }, + { + "epoch": 3.1690694626474443, + "grad_norm": 0.7777029275894165, + "learning_rate": 4.317151635641569e-05, + "loss": 0.3107, + "num_input_tokens_seen": 3238552, + "step": 12090 + }, + { + "epoch": 3.1703800786369594, + "grad_norm": 0.9358748197555542, + "learning_rate": 4.316365950164404e-05, + "loss": 0.4435, + "num_input_tokens_seen": 3239704, + "step": 12095 + }, + { + "epoch": 3.1716906946264745, + "grad_norm": 0.5120379328727722, + "learning_rate": 4.3155798845249827e-05, + "loss": 0.4967, + "num_input_tokens_seen": 3241256, + "step": 12100 + }, + { + "epoch": 3.1730013106159896, + "grad_norm": 0.8182370066642761, + "learning_rate": 4.3147934388878295e-05, + "loss": 0.3817, + "num_input_tokens_seen": 3242696, + "step": 12105 + }, + { + "epoch": 3.1743119266055047, + "grad_norm": 1.3555338382720947, + "learning_rate": 4.314006613417546e-05, + "loss": 0.4294, + "num_input_tokens_seen": 3244280, + "step": 12110 + }, + { + "epoch": 3.1756225425950197, + "grad_norm": 0.8264948725700378, + "learning_rate": 4.3132194082788116e-05, + "loss": 0.2806, + "num_input_tokens_seen": 3245608, + "step": 12115 + }, + { + "epoch": 3.176933158584535, + "grad_norm": 1.0875308513641357, + "learning_rate": 4.312431823636388e-05, + "loss": 0.3158, + "num_input_tokens_seen": 3246904, + "step": 12120 + }, + { + "epoch": 3.17824377457405, + "grad_norm": 0.8741095066070557, + "learning_rate": 4.311643859655116e-05, + "loss": 0.3654, + "num_input_tokens_seen": 3248360, + "step": 12125 + }, + { + "epoch": 3.179554390563565, + "grad_norm": 0.7903116941452026, + "learning_rate": 4.3108555164999144e-05, + "loss": 0.4307, + "num_input_tokens_seen": 3249560, + "step": 12130 + }, + { + "epoch": 3.18086500655308, + "grad_norm": 1.8177214860916138, + "learning_rate": 4.310066794335782e-05, + "loss": 0.4303, + "num_input_tokens_seen": 3250792, + "step": 12135 + }, + { + "epoch": 3.1821756225425952, + "grad_norm": 0.6624869108200073, + "learning_rate": 4.309277693327797e-05, + "loss": 0.3017, + "num_input_tokens_seen": 3252120, + "step": 12140 + }, + { + "epoch": 3.18348623853211, + "grad_norm": 0.9711213111877441, + "learning_rate": 4.308488213641116e-05, + "loss": 0.2989, + "num_input_tokens_seen": 3253432, + "step": 12145 + }, + { + "epoch": 3.184796854521625, + "grad_norm": 1.4370126724243164, + "learning_rate": 4.3076983554409784e-05, + "loss": 0.3667, + "num_input_tokens_seen": 3254600, + "step": 12150 + }, + { + "epoch": 3.18610747051114, + "grad_norm": 1.0134024620056152, + "learning_rate": 4.306908118892698e-05, + "loss": 0.4297, + "num_input_tokens_seen": 3255704, + "step": 12155 + }, + { + "epoch": 3.187418086500655, + "grad_norm": 0.8244575262069702, + "learning_rate": 4.30611750416167e-05, + "loss": 0.3356, + "num_input_tokens_seen": 3256904, + "step": 12160 + }, + { + "epoch": 3.1887287024901703, + "grad_norm": 0.685359537601471, + "learning_rate": 4.305326511413369e-05, + "loss": 0.2908, + "num_input_tokens_seen": 3258184, + "step": 12165 + }, + { + "epoch": 3.1900393184796854, + "grad_norm": 0.3520451486110687, + "learning_rate": 4.304535140813349e-05, + "loss": 0.4185, + "num_input_tokens_seen": 3259432, + "step": 12170 + }, + { + "epoch": 3.1913499344692005, + "grad_norm": 1.2719545364379883, + "learning_rate": 4.303743392527242e-05, + "loss": 0.4594, + "num_input_tokens_seen": 3260616, + "step": 12175 + }, + { + "epoch": 3.1926605504587156, + "grad_norm": 0.8235510587692261, + "learning_rate": 4.302951266720759e-05, + "loss": 0.534, + "num_input_tokens_seen": 3262200, + "step": 12180 + }, + { + "epoch": 3.1939711664482306, + "grad_norm": 0.5503410696983337, + "learning_rate": 4.3021587635596914e-05, + "loss": 0.3532, + "num_input_tokens_seen": 3264136, + "step": 12185 + }, + { + "epoch": 3.1952817824377457, + "grad_norm": 2.2011005878448486, + "learning_rate": 4.301365883209909e-05, + "loss": 0.5683, + "num_input_tokens_seen": 3265112, + "step": 12190 + }, + { + "epoch": 3.196592398427261, + "grad_norm": 1.4408937692642212, + "learning_rate": 4.300572625837359e-05, + "loss": 0.4884, + "num_input_tokens_seen": 3268472, + "step": 12195 + }, + { + "epoch": 3.197903014416776, + "grad_norm": 0.7999950051307678, + "learning_rate": 4.29977899160807e-05, + "loss": 0.5123, + "num_input_tokens_seen": 3270168, + "step": 12200 + }, + { + "epoch": 3.199213630406291, + "grad_norm": 0.625616192817688, + "learning_rate": 4.298984980688148e-05, + "loss": 0.304, + "num_input_tokens_seen": 3271752, + "step": 12205 + }, + { + "epoch": 3.200524246395806, + "grad_norm": 2.840463638305664, + "learning_rate": 4.298190593243778e-05, + "loss": 0.3346, + "num_input_tokens_seen": 3272888, + "step": 12210 + }, + { + "epoch": 3.2018348623853212, + "grad_norm": 0.6932035684585571, + "learning_rate": 4.297395829441223e-05, + "loss": 0.528, + "num_input_tokens_seen": 3273960, + "step": 12215 + }, + { + "epoch": 3.2031454783748363, + "grad_norm": 0.9588943123817444, + "learning_rate": 4.2966006894468274e-05, + "loss": 0.478, + "num_input_tokens_seen": 3275240, + "step": 12220 + }, + { + "epoch": 3.2044560943643514, + "grad_norm": 0.6763355135917664, + "learning_rate": 4.295805173427012e-05, + "loss": 0.3644, + "num_input_tokens_seen": 3277096, + "step": 12225 + }, + { + "epoch": 3.2057667103538665, + "grad_norm": 1.1704293489456177, + "learning_rate": 4.295009281548276e-05, + "loss": 0.3813, + "num_input_tokens_seen": 3278744, + "step": 12230 + }, + { + "epoch": 3.2070773263433816, + "grad_norm": 0.8622516393661499, + "learning_rate": 4.294213013977199e-05, + "loss": 0.337, + "num_input_tokens_seen": 3280616, + "step": 12235 + }, + { + "epoch": 3.2083879423328963, + "grad_norm": 1.3388150930404663, + "learning_rate": 4.2934163708804385e-05, + "loss": 0.4558, + "num_input_tokens_seen": 3282008, + "step": 12240 + }, + { + "epoch": 3.2096985583224114, + "grad_norm": 1.4439494609832764, + "learning_rate": 4.29261935242473e-05, + "loss": 0.4207, + "num_input_tokens_seen": 3283288, + "step": 12245 + }, + { + "epoch": 3.2110091743119265, + "grad_norm": 0.754450798034668, + "learning_rate": 4.291821958776889e-05, + "loss": 0.3631, + "num_input_tokens_seen": 3284696, + "step": 12250 + }, + { + "epoch": 3.2123197903014415, + "grad_norm": 0.6458016037940979, + "learning_rate": 4.291024190103808e-05, + "loss": 0.4117, + "num_input_tokens_seen": 3285880, + "step": 12255 + }, + { + "epoch": 3.2136304062909566, + "grad_norm": 0.625478208065033, + "learning_rate": 4.290226046572459e-05, + "loss": 0.5161, + "num_input_tokens_seen": 3287112, + "step": 12260 + }, + { + "epoch": 3.2149410222804717, + "grad_norm": 1.1239184141159058, + "learning_rate": 4.289427528349891e-05, + "loss": 0.4489, + "num_input_tokens_seen": 3288792, + "step": 12265 + }, + { + "epoch": 3.216251638269987, + "grad_norm": 0.9425492286682129, + "learning_rate": 4.2886286356032334e-05, + "loss": 0.3408, + "num_input_tokens_seen": 3290104, + "step": 12270 + }, + { + "epoch": 3.217562254259502, + "grad_norm": 0.7788508534431458, + "learning_rate": 4.287829368499693e-05, + "loss": 0.4164, + "num_input_tokens_seen": 3291304, + "step": 12275 + }, + { + "epoch": 3.218872870249017, + "grad_norm": 1.6116572618484497, + "learning_rate": 4.2870297272065546e-05, + "loss": 0.4914, + "num_input_tokens_seen": 3292600, + "step": 12280 + }, + { + "epoch": 3.220183486238532, + "grad_norm": 1.0444190502166748, + "learning_rate": 4.286229711891182e-05, + "loss": 0.5052, + "num_input_tokens_seen": 3293672, + "step": 12285 + }, + { + "epoch": 3.221494102228047, + "grad_norm": 1.0833543539047241, + "learning_rate": 4.285429322721017e-05, + "loss": 0.4786, + "num_input_tokens_seen": 3295192, + "step": 12290 + }, + { + "epoch": 3.2228047182175623, + "grad_norm": 0.5592139363288879, + "learning_rate": 4.284628559863581e-05, + "loss": 0.3945, + "num_input_tokens_seen": 3296216, + "step": 12295 + }, + { + "epoch": 3.2241153342070774, + "grad_norm": 0.8252426385879517, + "learning_rate": 4.2838274234864694e-05, + "loss": 0.457, + "num_input_tokens_seen": 3297464, + "step": 12300 + }, + { + "epoch": 3.2254259501965925, + "grad_norm": 1.0642890930175781, + "learning_rate": 4.283025913757361e-05, + "loss": 0.5819, + "num_input_tokens_seen": 3298744, + "step": 12305 + }, + { + "epoch": 3.2267365661861076, + "grad_norm": 0.677105724811554, + "learning_rate": 4.2822240308440096e-05, + "loss": 0.2139, + "num_input_tokens_seen": 3299848, + "step": 12310 + }, + { + "epoch": 3.2280471821756227, + "grad_norm": 0.7260911464691162, + "learning_rate": 4.281421774914248e-05, + "loss": 0.2597, + "num_input_tokens_seen": 3301432, + "step": 12315 + }, + { + "epoch": 3.229357798165138, + "grad_norm": 1.2075631618499756, + "learning_rate": 4.280619146135988e-05, + "loss": 0.471, + "num_input_tokens_seen": 3303176, + "step": 12320 + }, + { + "epoch": 3.230668414154653, + "grad_norm": 1.1385798454284668, + "learning_rate": 4.2798161446772156e-05, + "loss": 0.4342, + "num_input_tokens_seen": 3304824, + "step": 12325 + }, + { + "epoch": 3.231979030144168, + "grad_norm": 1.3813751935958862, + "learning_rate": 4.279012770706e-05, + "loss": 0.3074, + "num_input_tokens_seen": 3306312, + "step": 12330 + }, + { + "epoch": 3.2332896461336826, + "grad_norm": 1.2623279094696045, + "learning_rate": 4.278209024390486e-05, + "loss": 0.4195, + "num_input_tokens_seen": 3307864, + "step": 12335 + }, + { + "epoch": 3.2346002621231977, + "grad_norm": 1.9289064407348633, + "learning_rate": 4.2774049058988955e-05, + "loss": 0.4659, + "num_input_tokens_seen": 3309304, + "step": 12340 + }, + { + "epoch": 3.235910878112713, + "grad_norm": 1.011432409286499, + "learning_rate": 4.276600415399529e-05, + "loss": 0.3406, + "num_input_tokens_seen": 3310568, + "step": 12345 + }, + { + "epoch": 3.237221494102228, + "grad_norm": 1.3131110668182373, + "learning_rate": 4.275795553060765e-05, + "loss": 0.3521, + "num_input_tokens_seen": 3312024, + "step": 12350 + }, + { + "epoch": 3.238532110091743, + "grad_norm": 0.9449948668479919, + "learning_rate": 4.274990319051061e-05, + "loss": 0.4076, + "num_input_tokens_seen": 3313336, + "step": 12355 + }, + { + "epoch": 3.239842726081258, + "grad_norm": 1.5686594247817993, + "learning_rate": 4.2741847135389495e-05, + "loss": 0.5152, + "num_input_tokens_seen": 3314616, + "step": 12360 + }, + { + "epoch": 3.241153342070773, + "grad_norm": 1.012629508972168, + "learning_rate": 4.2733787366930436e-05, + "loss": 0.4068, + "num_input_tokens_seen": 3315864, + "step": 12365 + }, + { + "epoch": 3.2424639580602883, + "grad_norm": 1.1001042127609253, + "learning_rate": 4.2725723886820314e-05, + "loss": 0.5671, + "num_input_tokens_seen": 3317144, + "step": 12370 + }, + { + "epoch": 3.2437745740498034, + "grad_norm": 0.5010740756988525, + "learning_rate": 4.271765669674682e-05, + "loss": 0.3519, + "num_input_tokens_seen": 3318728, + "step": 12375 + }, + { + "epoch": 3.2450851900393185, + "grad_norm": 0.5056045055389404, + "learning_rate": 4.270958579839839e-05, + "loss": 0.3144, + "num_input_tokens_seen": 3320008, + "step": 12380 + }, + { + "epoch": 3.2463958060288336, + "grad_norm": 0.6583124399185181, + "learning_rate": 4.270151119346424e-05, + "loss": 0.4802, + "num_input_tokens_seen": 3321496, + "step": 12385 + }, + { + "epoch": 3.2477064220183487, + "grad_norm": 1.696530818939209, + "learning_rate": 4.26934328836344e-05, + "loss": 0.4705, + "num_input_tokens_seen": 3322632, + "step": 12390 + }, + { + "epoch": 3.249017038007864, + "grad_norm": 0.7185325622558594, + "learning_rate": 4.268535087059962e-05, + "loss": 0.2731, + "num_input_tokens_seen": 3324168, + "step": 12395 + }, + { + "epoch": 3.250327653997379, + "grad_norm": 0.9626126289367676, + "learning_rate": 4.267726515605146e-05, + "loss": 0.3962, + "num_input_tokens_seen": 3325592, + "step": 12400 + }, + { + "epoch": 3.251638269986894, + "grad_norm": 1.9506617784500122, + "learning_rate": 4.2669175741682255e-05, + "loss": 0.3931, + "num_input_tokens_seen": 3326920, + "step": 12405 + }, + { + "epoch": 3.252948885976409, + "grad_norm": 1.3244434595108032, + "learning_rate": 4.266108262918509e-05, + "loss": 0.3757, + "num_input_tokens_seen": 3327912, + "step": 12410 + }, + { + "epoch": 3.254259501965924, + "grad_norm": 1.3413031101226807, + "learning_rate": 4.265298582025385e-05, + "loss": 0.6686, + "num_input_tokens_seen": 3329352, + "step": 12415 + }, + { + "epoch": 3.255570117955439, + "grad_norm": 1.4588861465454102, + "learning_rate": 4.264488531658317e-05, + "loss": 0.4976, + "num_input_tokens_seen": 3330456, + "step": 12420 + }, + { + "epoch": 3.2568807339449544, + "grad_norm": 1.1064460277557373, + "learning_rate": 4.26367811198685e-05, + "loss": 0.3666, + "num_input_tokens_seen": 3331752, + "step": 12425 + }, + { + "epoch": 3.258191349934469, + "grad_norm": 0.6476446390151978, + "learning_rate": 4.2628673231806e-05, + "loss": 0.4355, + "num_input_tokens_seen": 3333128, + "step": 12430 + }, + { + "epoch": 3.259501965923984, + "grad_norm": 0.841576874256134, + "learning_rate": 4.262056165409266e-05, + "loss": 0.402, + "num_input_tokens_seen": 3334200, + "step": 12435 + }, + { + "epoch": 3.260812581913499, + "grad_norm": 0.8978348970413208, + "learning_rate": 4.261244638842621e-05, + "loss": 0.2202, + "num_input_tokens_seen": 3335496, + "step": 12440 + }, + { + "epoch": 3.2621231979030143, + "grad_norm": 2.6753599643707275, + "learning_rate": 4.260432743650516e-05, + "loss": 0.6634, + "num_input_tokens_seen": 3336488, + "step": 12445 + }, + { + "epoch": 3.2634338138925294, + "grad_norm": 0.7570977807044983, + "learning_rate": 4.25962048000288e-05, + "loss": 0.4814, + "num_input_tokens_seen": 3337832, + "step": 12450 + }, + { + "epoch": 3.2647444298820445, + "grad_norm": 2.6115469932556152, + "learning_rate": 4.258807848069718e-05, + "loss": 0.5547, + "num_input_tokens_seen": 3339384, + "step": 12455 + }, + { + "epoch": 3.2660550458715596, + "grad_norm": 1.4924023151397705, + "learning_rate": 4.257994848021112e-05, + "loss": 0.5006, + "num_input_tokens_seen": 3340904, + "step": 12460 + }, + { + "epoch": 3.2673656618610747, + "grad_norm": 1.0342450141906738, + "learning_rate": 4.257181480027222e-05, + "loss": 0.5016, + "num_input_tokens_seen": 3342296, + "step": 12465 + }, + { + "epoch": 3.26867627785059, + "grad_norm": 1.1815825700759888, + "learning_rate": 4.256367744258285e-05, + "loss": 0.7109, + "num_input_tokens_seen": 3343336, + "step": 12470 + }, + { + "epoch": 3.269986893840105, + "grad_norm": 1.2234158515930176, + "learning_rate": 4.255553640884613e-05, + "loss": 0.2901, + "num_input_tokens_seen": 3344472, + "step": 12475 + }, + { + "epoch": 3.27129750982962, + "grad_norm": 0.996877133846283, + "learning_rate": 4.254739170076597e-05, + "loss": 0.3933, + "num_input_tokens_seen": 3345720, + "step": 12480 + }, + { + "epoch": 3.272608125819135, + "grad_norm": 2.0490939617156982, + "learning_rate": 4.253924332004706e-05, + "loss": 0.4161, + "num_input_tokens_seen": 3346920, + "step": 12485 + }, + { + "epoch": 3.27391874180865, + "grad_norm": 0.9629819989204407, + "learning_rate": 4.2531091268394816e-05, + "loss": 0.3644, + "num_input_tokens_seen": 3348232, + "step": 12490 + }, + { + "epoch": 3.2752293577981653, + "grad_norm": 1.8460094928741455, + "learning_rate": 4.252293554751547e-05, + "loss": 0.3619, + "num_input_tokens_seen": 3349400, + "step": 12495 + }, + { + "epoch": 3.2765399737876804, + "grad_norm": 1.2972536087036133, + "learning_rate": 4.251477615911599e-05, + "loss": 0.5961, + "num_input_tokens_seen": 3350552, + "step": 12500 + }, + { + "epoch": 3.2778505897771955, + "grad_norm": 3.8569271564483643, + "learning_rate": 4.250661310490412e-05, + "loss": 0.3668, + "num_input_tokens_seen": 3351832, + "step": 12505 + }, + { + "epoch": 3.2791612057667106, + "grad_norm": 0.35975879430770874, + "learning_rate": 4.249844638658838e-05, + "loss": 0.2603, + "num_input_tokens_seen": 3353512, + "step": 12510 + }, + { + "epoch": 3.280471821756225, + "grad_norm": 1.0130043029785156, + "learning_rate": 4.249027600587804e-05, + "loss": 0.3586, + "num_input_tokens_seen": 3354744, + "step": 12515 + }, + { + "epoch": 3.2817824377457407, + "grad_norm": 1.3204258680343628, + "learning_rate": 4.248210196448316e-05, + "loss": 0.314, + "num_input_tokens_seen": 3355960, + "step": 12520 + }, + { + "epoch": 3.2830930537352554, + "grad_norm": 1.9912434816360474, + "learning_rate": 4.247392426411454e-05, + "loss": 0.4184, + "num_input_tokens_seen": 3356952, + "step": 12525 + }, + { + "epoch": 3.2844036697247705, + "grad_norm": 0.5961123108863831, + "learning_rate": 4.246574290648376e-05, + "loss": 0.4711, + "num_input_tokens_seen": 3358312, + "step": 12530 + }, + { + "epoch": 3.2857142857142856, + "grad_norm": 0.7159281969070435, + "learning_rate": 4.245755789330318e-05, + "loss": 0.4218, + "num_input_tokens_seen": 3359912, + "step": 12535 + }, + { + "epoch": 3.2870249017038007, + "grad_norm": 1.4081355333328247, + "learning_rate": 4.2449369226285884e-05, + "loss": 0.4437, + "num_input_tokens_seen": 3360920, + "step": 12540 + }, + { + "epoch": 3.288335517693316, + "grad_norm": 1.2782914638519287, + "learning_rate": 4.2441176907145775e-05, + "loss": 0.3594, + "num_input_tokens_seen": 3362104, + "step": 12545 + }, + { + "epoch": 3.289646133682831, + "grad_norm": 2.5845189094543457, + "learning_rate": 4.243298093759746e-05, + "loss": 0.3186, + "num_input_tokens_seen": 3363288, + "step": 12550 + }, + { + "epoch": 3.290956749672346, + "grad_norm": 1.3217333555221558, + "learning_rate": 4.242478131935637e-05, + "loss": 0.3898, + "num_input_tokens_seen": 3364424, + "step": 12555 + }, + { + "epoch": 3.292267365661861, + "grad_norm": 1.1310023069381714, + "learning_rate": 4.241657805413865e-05, + "loss": 0.4653, + "num_input_tokens_seen": 3365736, + "step": 12560 + }, + { + "epoch": 3.293577981651376, + "grad_norm": 1.1784182786941528, + "learning_rate": 4.240837114366123e-05, + "loss": 0.3838, + "num_input_tokens_seen": 3366968, + "step": 12565 + }, + { + "epoch": 3.2948885976408913, + "grad_norm": 0.6646861433982849, + "learning_rate": 4.2400160589641825e-05, + "loss": 0.3126, + "num_input_tokens_seen": 3368216, + "step": 12570 + }, + { + "epoch": 3.2961992136304064, + "grad_norm": 0.9813991189002991, + "learning_rate": 4.239194639379887e-05, + "loss": 0.4153, + "num_input_tokens_seen": 3369464, + "step": 12575 + }, + { + "epoch": 3.2975098296199215, + "grad_norm": 1.0167714357376099, + "learning_rate": 4.238372855785158e-05, + "loss": 0.395, + "num_input_tokens_seen": 3370584, + "step": 12580 + }, + { + "epoch": 3.2988204456094365, + "grad_norm": 0.636497974395752, + "learning_rate": 4.2375507083519936e-05, + "loss": 0.4322, + "num_input_tokens_seen": 3371848, + "step": 12585 + }, + { + "epoch": 3.3001310615989516, + "grad_norm": 1.4070122241973877, + "learning_rate": 4.2367281972524685e-05, + "loss": 0.4081, + "num_input_tokens_seen": 3373128, + "step": 12590 + }, + { + "epoch": 3.3014416775884667, + "grad_norm": 1.351675271987915, + "learning_rate": 4.235905322658733e-05, + "loss": 0.447, + "num_input_tokens_seen": 3374296, + "step": 12595 + }, + { + "epoch": 3.302752293577982, + "grad_norm": 0.4671825170516968, + "learning_rate": 4.235082084743013e-05, + "loss": 0.2892, + "num_input_tokens_seen": 3375592, + "step": 12600 + }, + { + "epoch": 3.304062909567497, + "grad_norm": 0.7429592609405518, + "learning_rate": 4.23425848367761e-05, + "loss": 0.394, + "num_input_tokens_seen": 3377336, + "step": 12605 + }, + { + "epoch": 3.3053735255570116, + "grad_norm": 1.6682237386703491, + "learning_rate": 4.2334345196349036e-05, + "loss": 0.4431, + "num_input_tokens_seen": 3378440, + "step": 12610 + }, + { + "epoch": 3.3066841415465267, + "grad_norm": 0.8262801170349121, + "learning_rate": 4.2326101927873464e-05, + "loss": 0.4397, + "num_input_tokens_seen": 3380088, + "step": 12615 + }, + { + "epoch": 3.3079947575360418, + "grad_norm": 1.3378825187683105, + "learning_rate": 4.2317855033074706e-05, + "loss": 0.5059, + "num_input_tokens_seen": 3381208, + "step": 12620 + }, + { + "epoch": 3.309305373525557, + "grad_norm": 1.1124718189239502, + "learning_rate": 4.2309604513678804e-05, + "loss": 0.5732, + "num_input_tokens_seen": 3382568, + "step": 12625 + }, + { + "epoch": 3.310615989515072, + "grad_norm": 1.0476555824279785, + "learning_rate": 4.230135037141259e-05, + "loss": 0.7269, + "num_input_tokens_seen": 3383768, + "step": 12630 + }, + { + "epoch": 3.311926605504587, + "grad_norm": 1.9515618085861206, + "learning_rate": 4.229309260800364e-05, + "loss": 0.2816, + "num_input_tokens_seen": 3384712, + "step": 12635 + }, + { + "epoch": 3.313237221494102, + "grad_norm": 1.0226720571517944, + "learning_rate": 4.228483122518029e-05, + "loss": 0.4728, + "num_input_tokens_seen": 3386040, + "step": 12640 + }, + { + "epoch": 3.3145478374836173, + "grad_norm": 0.7853525876998901, + "learning_rate": 4.227656622467162e-05, + "loss": 0.385, + "num_input_tokens_seen": 3387208, + "step": 12645 + }, + { + "epoch": 3.3158584534731324, + "grad_norm": 0.9257129430770874, + "learning_rate": 4.2268297608207496e-05, + "loss": 0.4637, + "num_input_tokens_seen": 3388408, + "step": 12650 + }, + { + "epoch": 3.3171690694626474, + "grad_norm": 0.7748465538024902, + "learning_rate": 4.226002537751852e-05, + "loss": 0.4633, + "num_input_tokens_seen": 3389448, + "step": 12655 + }, + { + "epoch": 3.3184796854521625, + "grad_norm": 1.1285974979400635, + "learning_rate": 4.225174953433605e-05, + "loss": 0.4246, + "num_input_tokens_seen": 3390568, + "step": 12660 + }, + { + "epoch": 3.3197903014416776, + "grad_norm": 1.7338030338287354, + "learning_rate": 4.224347008039222e-05, + "loss": 0.3323, + "num_input_tokens_seen": 3391768, + "step": 12665 + }, + { + "epoch": 3.3211009174311927, + "grad_norm": 1.2607250213623047, + "learning_rate": 4.22351870174199e-05, + "loss": 0.4181, + "num_input_tokens_seen": 3393208, + "step": 12670 + }, + { + "epoch": 3.322411533420708, + "grad_norm": 0.8501483201980591, + "learning_rate": 4.222690034715271e-05, + "loss": 0.2595, + "num_input_tokens_seen": 3394248, + "step": 12675 + }, + { + "epoch": 3.323722149410223, + "grad_norm": 1.337336778640747, + "learning_rate": 4.2218610071325036e-05, + "loss": 0.4584, + "num_input_tokens_seen": 3396232, + "step": 12680 + }, + { + "epoch": 3.325032765399738, + "grad_norm": 0.6161709427833557, + "learning_rate": 4.221031619167204e-05, + "loss": 0.3484, + "num_input_tokens_seen": 3397432, + "step": 12685 + }, + { + "epoch": 3.326343381389253, + "grad_norm": 0.7921980619430542, + "learning_rate": 4.2202018709929595e-05, + "loss": 0.3366, + "num_input_tokens_seen": 3398760, + "step": 12690 + }, + { + "epoch": 3.327653997378768, + "grad_norm": 1.196533441543579, + "learning_rate": 4.219371762783436e-05, + "loss": 0.5251, + "num_input_tokens_seen": 3399848, + "step": 12695 + }, + { + "epoch": 3.3289646133682833, + "grad_norm": 0.7957841753959656, + "learning_rate": 4.218541294712373e-05, + "loss": 0.4646, + "num_input_tokens_seen": 3401240, + "step": 12700 + }, + { + "epoch": 3.330275229357798, + "grad_norm": 1.6759017705917358, + "learning_rate": 4.217710466953586e-05, + "loss": 1.0467, + "num_input_tokens_seen": 3402920, + "step": 12705 + }, + { + "epoch": 3.331585845347313, + "grad_norm": 0.35833507776260376, + "learning_rate": 4.2168792796809674e-05, + "loss": 0.3714, + "num_input_tokens_seen": 3404184, + "step": 12710 + }, + { + "epoch": 3.332896461336828, + "grad_norm": 0.5501057505607605, + "learning_rate": 4.216047733068481e-05, + "loss": 0.3518, + "num_input_tokens_seen": 3405576, + "step": 12715 + }, + { + "epoch": 3.3342070773263432, + "grad_norm": 1.2825895547866821, + "learning_rate": 4.21521582729017e-05, + "loss": 0.4728, + "num_input_tokens_seen": 3406712, + "step": 12720 + }, + { + "epoch": 3.3355176933158583, + "grad_norm": 2.230912923812866, + "learning_rate": 4.21438356252015e-05, + "loss": 0.4884, + "num_input_tokens_seen": 3407928, + "step": 12725 + }, + { + "epoch": 3.3368283093053734, + "grad_norm": 1.0902808904647827, + "learning_rate": 4.213550938932612e-05, + "loss": 0.5323, + "num_input_tokens_seen": 3409272, + "step": 12730 + }, + { + "epoch": 3.3381389252948885, + "grad_norm": 0.7702265381813049, + "learning_rate": 4.2127179567018235e-05, + "loss": 0.3373, + "num_input_tokens_seen": 3410568, + "step": 12735 + }, + { + "epoch": 3.3394495412844036, + "grad_norm": 1.2445486783981323, + "learning_rate": 4.211884616002126e-05, + "loss": 0.3495, + "num_input_tokens_seen": 3412248, + "step": 12740 + }, + { + "epoch": 3.3407601572739187, + "grad_norm": 0.6105205416679382, + "learning_rate": 4.211050917007937e-05, + "loss": 0.4353, + "num_input_tokens_seen": 3414104, + "step": 12745 + }, + { + "epoch": 3.342070773263434, + "grad_norm": 0.9381304979324341, + "learning_rate": 4.210216859893747e-05, + "loss": 0.3073, + "num_input_tokens_seen": 3415256, + "step": 12750 + }, + { + "epoch": 3.343381389252949, + "grad_norm": 4.204158782958984, + "learning_rate": 4.2093824448341226e-05, + "loss": 0.4557, + "num_input_tokens_seen": 3416264, + "step": 12755 + }, + { + "epoch": 3.344692005242464, + "grad_norm": 1.2046009302139282, + "learning_rate": 4.208547672003707e-05, + "loss": 0.3983, + "num_input_tokens_seen": 3417592, + "step": 12760 + }, + { + "epoch": 3.346002621231979, + "grad_norm": 0.5873125791549683, + "learning_rate": 4.207712541577215e-05, + "loss": 0.4268, + "num_input_tokens_seen": 3419032, + "step": 12765 + }, + { + "epoch": 3.347313237221494, + "grad_norm": 1.2176191806793213, + "learning_rate": 4.2068770537294395e-05, + "loss": 0.3815, + "num_input_tokens_seen": 3420104, + "step": 12770 + }, + { + "epoch": 3.3486238532110093, + "grad_norm": 0.9903748035430908, + "learning_rate": 4.2060412086352455e-05, + "loss": 0.3218, + "num_input_tokens_seen": 3421704, + "step": 12775 + }, + { + "epoch": 3.3499344692005244, + "grad_norm": 0.9516573548316956, + "learning_rate": 4.2052050064695746e-05, + "loss": 0.3517, + "num_input_tokens_seen": 3422904, + "step": 12780 + }, + { + "epoch": 3.3512450851900395, + "grad_norm": 0.748820424079895, + "learning_rate": 4.204368447407442e-05, + "loss": 0.3044, + "num_input_tokens_seen": 3424456, + "step": 12785 + }, + { + "epoch": 3.352555701179554, + "grad_norm": 0.4932038187980652, + "learning_rate": 4.203531531623938e-05, + "loss": 0.2629, + "num_input_tokens_seen": 3425576, + "step": 12790 + }, + { + "epoch": 3.3538663171690697, + "grad_norm": 0.7182132601737976, + "learning_rate": 4.2026942592942285e-05, + "loss": 0.3681, + "num_input_tokens_seen": 3426920, + "step": 12795 + }, + { + "epoch": 3.3551769331585843, + "grad_norm": 0.7114022970199585, + "learning_rate": 4.2018566305935525e-05, + "loss": 0.4257, + "num_input_tokens_seen": 3428504, + "step": 12800 + }, + { + "epoch": 3.3564875491480994, + "grad_norm": 0.6782711148262024, + "learning_rate": 4.201018645697224e-05, + "loss": 0.2916, + "num_input_tokens_seen": 3429416, + "step": 12805 + }, + { + "epoch": 3.3577981651376145, + "grad_norm": 1.1145846843719482, + "learning_rate": 4.200180304780632e-05, + "loss": 0.3275, + "num_input_tokens_seen": 3430776, + "step": 12810 + }, + { + "epoch": 3.3591087811271296, + "grad_norm": 1.5149401426315308, + "learning_rate": 4.1993416080192404e-05, + "loss": 0.5434, + "num_input_tokens_seen": 3431832, + "step": 12815 + }, + { + "epoch": 3.3604193971166447, + "grad_norm": 0.8090846538543701, + "learning_rate": 4.198502555588586e-05, + "loss": 0.5209, + "num_input_tokens_seen": 3432904, + "step": 12820 + }, + { + "epoch": 3.36173001310616, + "grad_norm": 1.2932729721069336, + "learning_rate": 4.197663147664281e-05, + "loss": 0.342, + "num_input_tokens_seen": 3433976, + "step": 12825 + }, + { + "epoch": 3.363040629095675, + "grad_norm": 1.0243442058563232, + "learning_rate": 4.196823384422014e-05, + "loss": 0.3664, + "num_input_tokens_seen": 3435320, + "step": 12830 + }, + { + "epoch": 3.36435124508519, + "grad_norm": 0.5509526133537292, + "learning_rate": 4.195983266037543e-05, + "loss": 0.3759, + "num_input_tokens_seen": 3436552, + "step": 12835 + }, + { + "epoch": 3.365661861074705, + "grad_norm": 0.8808721899986267, + "learning_rate": 4.1951427926867056e-05, + "loss": 0.3533, + "num_input_tokens_seen": 3437656, + "step": 12840 + }, + { + "epoch": 3.36697247706422, + "grad_norm": 3.634037733078003, + "learning_rate": 4.1943019645454115e-05, + "loss": 0.4454, + "num_input_tokens_seen": 3438616, + "step": 12845 + }, + { + "epoch": 3.3682830930537353, + "grad_norm": 0.8489037156105042, + "learning_rate": 4.193460781789642e-05, + "loss": 0.4485, + "num_input_tokens_seen": 3439848, + "step": 12850 + }, + { + "epoch": 3.3695937090432504, + "grad_norm": 0.5105524063110352, + "learning_rate": 4.192619244595458e-05, + "loss": 0.4078, + "num_input_tokens_seen": 3441176, + "step": 12855 + }, + { + "epoch": 3.3709043250327655, + "grad_norm": 0.9424646496772766, + "learning_rate": 4.191777353138991e-05, + "loss": 0.4583, + "num_input_tokens_seen": 3442696, + "step": 12860 + }, + { + "epoch": 3.3722149410222806, + "grad_norm": 0.9490683674812317, + "learning_rate": 4.1909351075964464e-05, + "loss": 0.3703, + "num_input_tokens_seen": 3443832, + "step": 12865 + }, + { + "epoch": 3.3735255570117957, + "grad_norm": 0.5251302719116211, + "learning_rate": 4.190092508144106e-05, + "loss": 0.3972, + "num_input_tokens_seen": 3444952, + "step": 12870 + }, + { + "epoch": 3.374836173001311, + "grad_norm": 0.7611986398696899, + "learning_rate": 4.189249554958324e-05, + "loss": 0.3608, + "num_input_tokens_seen": 3446232, + "step": 12875 + }, + { + "epoch": 3.376146788990826, + "grad_norm": 0.6694284081459045, + "learning_rate": 4.188406248215528e-05, + "loss": 0.3526, + "num_input_tokens_seen": 3447528, + "step": 12880 + }, + { + "epoch": 3.3774574049803405, + "grad_norm": 1.0024490356445312, + "learning_rate": 4.1875625880922225e-05, + "loss": 0.739, + "num_input_tokens_seen": 3449352, + "step": 12885 + }, + { + "epoch": 3.378768020969856, + "grad_norm": 0.8839256763458252, + "learning_rate": 4.186718574764983e-05, + "loss": 0.5389, + "num_input_tokens_seen": 3450536, + "step": 12890 + }, + { + "epoch": 3.3800786369593707, + "grad_norm": 0.9514108300209045, + "learning_rate": 4.18587420841046e-05, + "loss": 0.3815, + "num_input_tokens_seen": 3451688, + "step": 12895 + }, + { + "epoch": 3.381389252948886, + "grad_norm": 1.040200114250183, + "learning_rate": 4.185029489205379e-05, + "loss": 0.408, + "num_input_tokens_seen": 3452760, + "step": 12900 + }, + { + "epoch": 3.382699868938401, + "grad_norm": 0.6283470988273621, + "learning_rate": 4.1841844173265357e-05, + "loss": 0.3342, + "num_input_tokens_seen": 3454712, + "step": 12905 + }, + { + "epoch": 3.384010484927916, + "grad_norm": 1.0171481370925903, + "learning_rate": 4.183338992950805e-05, + "loss": 0.4669, + "num_input_tokens_seen": 3456632, + "step": 12910 + }, + { + "epoch": 3.385321100917431, + "grad_norm": 1.2171871662139893, + "learning_rate": 4.182493216255132e-05, + "loss": 0.3798, + "num_input_tokens_seen": 3457992, + "step": 12915 + }, + { + "epoch": 3.386631716906946, + "grad_norm": 0.84587162733078, + "learning_rate": 4.181647087416536e-05, + "loss": 0.2629, + "num_input_tokens_seen": 3459224, + "step": 12920 + }, + { + "epoch": 3.3879423328964613, + "grad_norm": 0.9317967295646667, + "learning_rate": 4.18080060661211e-05, + "loss": 0.3652, + "num_input_tokens_seen": 3460680, + "step": 12925 + }, + { + "epoch": 3.3892529488859764, + "grad_norm": 1.0676826238632202, + "learning_rate": 4.179953774019021e-05, + "loss": 0.4239, + "num_input_tokens_seen": 3462520, + "step": 12930 + }, + { + "epoch": 3.3905635648754915, + "grad_norm": 0.8847527503967285, + "learning_rate": 4.1791065898145095e-05, + "loss": 0.3743, + "num_input_tokens_seen": 3463768, + "step": 12935 + }, + { + "epoch": 3.3918741808650066, + "grad_norm": 0.6390673518180847, + "learning_rate": 4.178259054175891e-05, + "loss": 0.3534, + "num_input_tokens_seen": 3465032, + "step": 12940 + }, + { + "epoch": 3.3931847968545217, + "grad_norm": 1.0591048002243042, + "learning_rate": 4.177411167280551e-05, + "loss": 0.3675, + "num_input_tokens_seen": 3466520, + "step": 12945 + }, + { + "epoch": 3.3944954128440368, + "grad_norm": 1.5606651306152344, + "learning_rate": 4.176562929305953e-05, + "loss": 0.5296, + "num_input_tokens_seen": 3467800, + "step": 12950 + }, + { + "epoch": 3.395806028833552, + "grad_norm": 1.179410696029663, + "learning_rate": 4.1757143404296305e-05, + "loss": 0.3691, + "num_input_tokens_seen": 3468840, + "step": 12955 + }, + { + "epoch": 3.397116644823067, + "grad_norm": 0.6115478277206421, + "learning_rate": 4.1748654008291914e-05, + "loss": 0.5245, + "num_input_tokens_seen": 3469912, + "step": 12960 + }, + { + "epoch": 3.398427260812582, + "grad_norm": 0.8376467823982239, + "learning_rate": 4.1740161106823185e-05, + "loss": 0.3527, + "num_input_tokens_seen": 3471736, + "step": 12965 + }, + { + "epoch": 3.399737876802097, + "grad_norm": 0.5665726661682129, + "learning_rate": 4.173166470166766e-05, + "loss": 0.3505, + "num_input_tokens_seen": 3473304, + "step": 12970 + }, + { + "epoch": 3.4010484927916123, + "grad_norm": 0.9440449476242065, + "learning_rate": 4.172316479460362e-05, + "loss": 0.4045, + "num_input_tokens_seen": 3474552, + "step": 12975 + }, + { + "epoch": 3.402359108781127, + "grad_norm": 0.7133580446243286, + "learning_rate": 4.171466138741009e-05, + "loss": 0.435, + "num_input_tokens_seen": 3475864, + "step": 12980 + }, + { + "epoch": 3.4036697247706424, + "grad_norm": 0.30543097853660583, + "learning_rate": 4.170615448186681e-05, + "loss": 0.3328, + "num_input_tokens_seen": 3477288, + "step": 12985 + }, + { + "epoch": 3.404980340760157, + "grad_norm": 0.6894322633743286, + "learning_rate": 4.169764407975426e-05, + "loss": 0.3456, + "num_input_tokens_seen": 3478712, + "step": 12990 + }, + { + "epoch": 3.406290956749672, + "grad_norm": 0.978475034236908, + "learning_rate": 4.168913018285366e-05, + "loss": 0.6317, + "num_input_tokens_seen": 3480040, + "step": 12995 + }, + { + "epoch": 3.4076015727391873, + "grad_norm": 0.9230387806892395, + "learning_rate": 4.168061279294694e-05, + "loss": 0.6444, + "num_input_tokens_seen": 3481864, + "step": 13000 + }, + { + "epoch": 3.4089121887287024, + "grad_norm": 0.83223956823349, + "learning_rate": 4.16720919118168e-05, + "loss": 0.3922, + "num_input_tokens_seen": 3483224, + "step": 13005 + }, + { + "epoch": 3.4102228047182175, + "grad_norm": 0.3571876287460327, + "learning_rate": 4.1663567541246615e-05, + "loss": 0.3439, + "num_input_tokens_seen": 3484728, + "step": 13010 + }, + { + "epoch": 3.4115334207077326, + "grad_norm": 0.6297502517700195, + "learning_rate": 4.165503968302055e-05, + "loss": 0.4969, + "num_input_tokens_seen": 3486376, + "step": 13015 + }, + { + "epoch": 3.4128440366972477, + "grad_norm": 1.1032291650772095, + "learning_rate": 4.1646508338923446e-05, + "loss": 0.5586, + "num_input_tokens_seen": 3487640, + "step": 13020 + }, + { + "epoch": 3.4141546526867628, + "grad_norm": 0.6455807089805603, + "learning_rate": 4.1637973510740914e-05, + "loss": 0.3994, + "num_input_tokens_seen": 3489160, + "step": 13025 + }, + { + "epoch": 3.415465268676278, + "grad_norm": 0.5640188455581665, + "learning_rate": 4.162943520025927e-05, + "loss": 0.3836, + "num_input_tokens_seen": 3490888, + "step": 13030 + }, + { + "epoch": 3.416775884665793, + "grad_norm": 0.9559222459793091, + "learning_rate": 4.162089340926558e-05, + "loss": 0.5215, + "num_input_tokens_seen": 3492344, + "step": 13035 + }, + { + "epoch": 3.418086500655308, + "grad_norm": 0.5104150772094727, + "learning_rate": 4.161234813954761e-05, + "loss": 0.4906, + "num_input_tokens_seen": 3493384, + "step": 13040 + }, + { + "epoch": 3.419397116644823, + "grad_norm": 0.5452298521995544, + "learning_rate": 4.160379939289388e-05, + "loss": 0.3092, + "num_input_tokens_seen": 3494760, + "step": 13045 + }, + { + "epoch": 3.4207077326343382, + "grad_norm": 1.1372517347335815, + "learning_rate": 4.1595247171093635e-05, + "loss": 0.3946, + "num_input_tokens_seen": 3495896, + "step": 13050 + }, + { + "epoch": 3.4220183486238533, + "grad_norm": 0.606529951095581, + "learning_rate": 4.158669147593682e-05, + "loss": 0.4925, + "num_input_tokens_seen": 3497160, + "step": 13055 + }, + { + "epoch": 3.4233289646133684, + "grad_norm": 2.3019323348999023, + "learning_rate": 4.157813230921414e-05, + "loss": 0.2923, + "num_input_tokens_seen": 3498472, + "step": 13060 + }, + { + "epoch": 3.4246395806028835, + "grad_norm": 1.404129981994629, + "learning_rate": 4.156956967271702e-05, + "loss": 0.514, + "num_input_tokens_seen": 3499688, + "step": 13065 + }, + { + "epoch": 3.4259501965923986, + "grad_norm": 1.1919755935668945, + "learning_rate": 4.156100356823759e-05, + "loss": 0.2567, + "num_input_tokens_seen": 3500984, + "step": 13070 + }, + { + "epoch": 3.4272608125819133, + "grad_norm": 1.2048927545547485, + "learning_rate": 4.1552433997568736e-05, + "loss": 0.817, + "num_input_tokens_seen": 3502184, + "step": 13075 + }, + { + "epoch": 3.4285714285714284, + "grad_norm": 0.24907474219799042, + "learning_rate": 4.154386096250404e-05, + "loss": 0.3549, + "num_input_tokens_seen": 3503640, + "step": 13080 + }, + { + "epoch": 3.4298820445609435, + "grad_norm": 1.4551626443862915, + "learning_rate": 4.153528446483783e-05, + "loss": 0.396, + "num_input_tokens_seen": 3505032, + "step": 13085 + }, + { + "epoch": 3.4311926605504586, + "grad_norm": 0.9869606494903564, + "learning_rate": 4.1526704506365153e-05, + "loss": 0.4259, + "num_input_tokens_seen": 3506088, + "step": 13090 + }, + { + "epoch": 3.4325032765399737, + "grad_norm": 0.40871936082839966, + "learning_rate": 4.1518121088881785e-05, + "loss": 0.5663, + "num_input_tokens_seen": 3507048, + "step": 13095 + }, + { + "epoch": 3.4338138925294888, + "grad_norm": 0.5569818019866943, + "learning_rate": 4.1509534214184206e-05, + "loss": 0.3425, + "num_input_tokens_seen": 3508296, + "step": 13100 + }, + { + "epoch": 3.435124508519004, + "grad_norm": 1.989530086517334, + "learning_rate": 4.1500943884069644e-05, + "loss": 0.7316, + "num_input_tokens_seen": 3509496, + "step": 13105 + }, + { + "epoch": 3.436435124508519, + "grad_norm": 0.5127525329589844, + "learning_rate": 4.149235010033604e-05, + "loss": 0.3824, + "num_input_tokens_seen": 3511080, + "step": 13110 + }, + { + "epoch": 3.437745740498034, + "grad_norm": 0.6760953068733215, + "learning_rate": 4.148375286478206e-05, + "loss": 0.3905, + "num_input_tokens_seen": 3512440, + "step": 13115 + }, + { + "epoch": 3.439056356487549, + "grad_norm": 0.8114026188850403, + "learning_rate": 4.147515217920709e-05, + "loss": 0.3584, + "num_input_tokens_seen": 3513720, + "step": 13120 + }, + { + "epoch": 3.4403669724770642, + "grad_norm": 1.7252346277236938, + "learning_rate": 4.146654804541122e-05, + "loss": 0.4368, + "num_input_tokens_seen": 3515240, + "step": 13125 + }, + { + "epoch": 3.4416775884665793, + "grad_norm": 0.5304789543151855, + "learning_rate": 4.145794046519531e-05, + "loss": 0.34, + "num_input_tokens_seen": 3516424, + "step": 13130 + }, + { + "epoch": 3.4429882044560944, + "grad_norm": 1.1799098253250122, + "learning_rate": 4.14493294403609e-05, + "loss": 0.5692, + "num_input_tokens_seen": 3517592, + "step": 13135 + }, + { + "epoch": 3.4442988204456095, + "grad_norm": 1.4684995412826538, + "learning_rate": 4.144071497271025e-05, + "loss": 0.337, + "num_input_tokens_seen": 3518728, + "step": 13140 + }, + { + "epoch": 3.4456094364351246, + "grad_norm": 0.8649064898490906, + "learning_rate": 4.143209706404637e-05, + "loss": 0.3537, + "num_input_tokens_seen": 3519960, + "step": 13145 + }, + { + "epoch": 3.4469200524246397, + "grad_norm": 0.8173993229866028, + "learning_rate": 4.142347571617296e-05, + "loss": 0.3317, + "num_input_tokens_seen": 3521096, + "step": 13150 + }, + { + "epoch": 3.448230668414155, + "grad_norm": 0.4889248013496399, + "learning_rate": 4.1414850930894466e-05, + "loss": 0.4853, + "num_input_tokens_seen": 3522360, + "step": 13155 + }, + { + "epoch": 3.44954128440367, + "grad_norm": 1.0406500101089478, + "learning_rate": 4.1406222710016034e-05, + "loss": 0.2938, + "num_input_tokens_seen": 3523768, + "step": 13160 + }, + { + "epoch": 3.450851900393185, + "grad_norm": 1.3474293947219849, + "learning_rate": 4.1397591055343534e-05, + "loss": 0.4118, + "num_input_tokens_seen": 3525368, + "step": 13165 + }, + { + "epoch": 3.4521625163826997, + "grad_norm": 3.7605056762695312, + "learning_rate": 4.1388955968683565e-05, + "loss": 0.7666, + "num_input_tokens_seen": 3526664, + "step": 13170 + }, + { + "epoch": 3.4534731323722148, + "grad_norm": 0.9190518260002136, + "learning_rate": 4.1380317451843425e-05, + "loss": 0.4677, + "num_input_tokens_seen": 3528088, + "step": 13175 + }, + { + "epoch": 3.45478374836173, + "grad_norm": 0.6002042293548584, + "learning_rate": 4.137167550663115e-05, + "loss": 0.295, + "num_input_tokens_seen": 3529496, + "step": 13180 + }, + { + "epoch": 3.456094364351245, + "grad_norm": 1.1499582529067993, + "learning_rate": 4.136303013485547e-05, + "loss": 0.3824, + "num_input_tokens_seen": 3531720, + "step": 13185 + }, + { + "epoch": 3.45740498034076, + "grad_norm": 0.5197351574897766, + "learning_rate": 4.1354381338325864e-05, + "loss": 0.4613, + "num_input_tokens_seen": 3533032, + "step": 13190 + }, + { + "epoch": 3.458715596330275, + "grad_norm": 1.8657381534576416, + "learning_rate": 4.13457291188525e-05, + "loss": 0.5375, + "num_input_tokens_seen": 3534136, + "step": 13195 + }, + { + "epoch": 3.4600262123197902, + "grad_norm": 0.5369338989257812, + "learning_rate": 4.1337073478246275e-05, + "loss": 0.6634, + "num_input_tokens_seen": 3535272, + "step": 13200 + }, + { + "epoch": 3.4613368283093053, + "grad_norm": 0.5888460874557495, + "learning_rate": 4.1328414418318795e-05, + "loss": 0.3228, + "num_input_tokens_seen": 3536536, + "step": 13205 + }, + { + "epoch": 3.4626474442988204, + "grad_norm": 2.3663859367370605, + "learning_rate": 4.13197519408824e-05, + "loss": 0.3696, + "num_input_tokens_seen": 3537496, + "step": 13210 + }, + { + "epoch": 3.4639580602883355, + "grad_norm": 1.1704862117767334, + "learning_rate": 4.1311086047750114e-05, + "loss": 0.4494, + "num_input_tokens_seen": 3538696, + "step": 13215 + }, + { + "epoch": 3.4652686762778506, + "grad_norm": 0.9111568927764893, + "learning_rate": 4.13024167407357e-05, + "loss": 0.6978, + "num_input_tokens_seen": 3540184, + "step": 13220 + }, + { + "epoch": 3.4665792922673657, + "grad_norm": 1.1865134239196777, + "learning_rate": 4.129374402165363e-05, + "loss": 0.3214, + "num_input_tokens_seen": 3541576, + "step": 13225 + }, + { + "epoch": 3.467889908256881, + "grad_norm": 0.7712523341178894, + "learning_rate": 4.128506789231911e-05, + "loss": 0.3514, + "num_input_tokens_seen": 3544776, + "step": 13230 + }, + { + "epoch": 3.469200524246396, + "grad_norm": 1.2305012941360474, + "learning_rate": 4.1276388354548e-05, + "loss": 0.3034, + "num_input_tokens_seen": 3545976, + "step": 13235 + }, + { + "epoch": 3.470511140235911, + "grad_norm": 0.7585687041282654, + "learning_rate": 4.126770541015693e-05, + "loss": 0.3163, + "num_input_tokens_seen": 3547160, + "step": 13240 + }, + { + "epoch": 3.471821756225426, + "grad_norm": 0.8788462281227112, + "learning_rate": 4.1259019060963236e-05, + "loss": 0.3108, + "num_input_tokens_seen": 3548296, + "step": 13245 + }, + { + "epoch": 3.473132372214941, + "grad_norm": 0.8848581314086914, + "learning_rate": 4.125032930878493e-05, + "loss": 0.3291, + "num_input_tokens_seen": 3549624, + "step": 13250 + }, + { + "epoch": 3.474442988204456, + "grad_norm": 1.5737526416778564, + "learning_rate": 4.1241636155440785e-05, + "loss": 0.3871, + "num_input_tokens_seen": 3550824, + "step": 13255 + }, + { + "epoch": 3.4757536041939714, + "grad_norm": 0.5375485420227051, + "learning_rate": 4.123293960275027e-05, + "loss": 0.3628, + "num_input_tokens_seen": 3552232, + "step": 13260 + }, + { + "epoch": 3.477064220183486, + "grad_norm": 1.1147191524505615, + "learning_rate": 4.1224239652533526e-05, + "loss": 0.4636, + "num_input_tokens_seen": 3553512, + "step": 13265 + }, + { + "epoch": 3.478374836173001, + "grad_norm": 0.6619154810905457, + "learning_rate": 4.121553630661147e-05, + "loss": 0.323, + "num_input_tokens_seen": 3554728, + "step": 13270 + }, + { + "epoch": 3.4796854521625162, + "grad_norm": 1.3996131420135498, + "learning_rate": 4.120682956680567e-05, + "loss": 0.5968, + "num_input_tokens_seen": 3555800, + "step": 13275 + }, + { + "epoch": 3.4809960681520313, + "grad_norm": 0.41406095027923584, + "learning_rate": 4.1198119434938455e-05, + "loss": 0.4268, + "num_input_tokens_seen": 3557400, + "step": 13280 + }, + { + "epoch": 3.4823066841415464, + "grad_norm": 0.6059225797653198, + "learning_rate": 4.118940591283282e-05, + "loss": 0.3639, + "num_input_tokens_seen": 3558504, + "step": 13285 + }, + { + "epoch": 3.4836173001310615, + "grad_norm": 0.6336900591850281, + "learning_rate": 4.118068900231251e-05, + "loss": 0.4258, + "num_input_tokens_seen": 3559928, + "step": 13290 + }, + { + "epoch": 3.4849279161205766, + "grad_norm": 2.080683469772339, + "learning_rate": 4.117196870520195e-05, + "loss": 0.8284, + "num_input_tokens_seen": 3561320, + "step": 13295 + }, + { + "epoch": 3.4862385321100917, + "grad_norm": 0.9158234596252441, + "learning_rate": 4.116324502332628e-05, + "loss": 0.5668, + "num_input_tokens_seen": 3562200, + "step": 13300 + }, + { + "epoch": 3.487549148099607, + "grad_norm": 0.44817331433296204, + "learning_rate": 4.115451795851136e-05, + "loss": 0.5506, + "num_input_tokens_seen": 3563544, + "step": 13305 + }, + { + "epoch": 3.488859764089122, + "grad_norm": 0.4296325445175171, + "learning_rate": 4.1145787512583736e-05, + "loss": 0.4438, + "num_input_tokens_seen": 3565320, + "step": 13310 + }, + { + "epoch": 3.490170380078637, + "grad_norm": 0.762884795665741, + "learning_rate": 4.113705368737069e-05, + "loss": 0.4653, + "num_input_tokens_seen": 3567288, + "step": 13315 + }, + { + "epoch": 3.491480996068152, + "grad_norm": 1.3295753002166748, + "learning_rate": 4.112831648470018e-05, + "loss": 0.4601, + "num_input_tokens_seen": 3568616, + "step": 13320 + }, + { + "epoch": 3.492791612057667, + "grad_norm": 0.479034960269928, + "learning_rate": 4.111957590640091e-05, + "loss": 0.4228, + "num_input_tokens_seen": 3570184, + "step": 13325 + }, + { + "epoch": 3.4941022280471823, + "grad_norm": 4.578223705291748, + "learning_rate": 4.111083195430225e-05, + "loss": 0.5275, + "num_input_tokens_seen": 3572040, + "step": 13330 + }, + { + "epoch": 3.4954128440366974, + "grad_norm": 0.7135145664215088, + "learning_rate": 4.110208463023429e-05, + "loss": 0.4301, + "num_input_tokens_seen": 3573560, + "step": 13335 + }, + { + "epoch": 3.4967234600262125, + "grad_norm": 0.2692844867706299, + "learning_rate": 4.109333393602786e-05, + "loss": 0.4921, + "num_input_tokens_seen": 3575048, + "step": 13340 + }, + { + "epoch": 3.4980340760157276, + "grad_norm": 0.5761278867721558, + "learning_rate": 4.108457987351443e-05, + "loss": 0.4203, + "num_input_tokens_seen": 3576408, + "step": 13345 + }, + { + "epoch": 3.4993446920052422, + "grad_norm": 0.8932007551193237, + "learning_rate": 4.107582244452622e-05, + "loss": 0.2849, + "num_input_tokens_seen": 3577448, + "step": 13350 + }, + { + "epoch": 3.5006553079947578, + "grad_norm": 0.8993467092514038, + "learning_rate": 4.106706165089616e-05, + "loss": 0.3807, + "num_input_tokens_seen": 3579256, + "step": 13355 + }, + { + "epoch": 3.5009174311926605, + "eval_loss": 0.5102009177207947, + "eval_runtime": 16.8753, + "eval_samples_per_second": 50.251, + "eval_steps_per_second": 25.125, + "num_input_tokens_seen": 3579432, + "step": 13356 + }, + { + "epoch": 3.5019659239842724, + "grad_norm": 1.1965957880020142, + "learning_rate": 4.105829749445785e-05, + "loss": 0.4248, + "num_input_tokens_seen": 3580584, + "step": 13360 + }, + { + "epoch": 3.5032765399737875, + "grad_norm": 0.5478951930999756, + "learning_rate": 4.104952997704563e-05, + "loss": 0.3418, + "num_input_tokens_seen": 3582440, + "step": 13365 + }, + { + "epoch": 3.5045871559633026, + "grad_norm": 0.9369902610778809, + "learning_rate": 4.1040759100494506e-05, + "loss": 0.3379, + "num_input_tokens_seen": 3583496, + "step": 13370 + }, + { + "epoch": 3.5058977719528177, + "grad_norm": 0.6654024124145508, + "learning_rate": 4.103198486664023e-05, + "loss": 0.2127, + "num_input_tokens_seen": 3584856, + "step": 13375 + }, + { + "epoch": 3.507208387942333, + "grad_norm": 1.431066632270813, + "learning_rate": 4.1023207277319213e-05, + "loss": 0.3744, + "num_input_tokens_seen": 3586120, + "step": 13380 + }, + { + "epoch": 3.508519003931848, + "grad_norm": 0.8483555912971497, + "learning_rate": 4.10144263343686e-05, + "loss": 0.2028, + "num_input_tokens_seen": 3587688, + "step": 13385 + }, + { + "epoch": 3.509829619921363, + "grad_norm": 1.4566751718521118, + "learning_rate": 4.1005642039626235e-05, + "loss": 0.6037, + "num_input_tokens_seen": 3588888, + "step": 13390 + }, + { + "epoch": 3.511140235910878, + "grad_norm": 1.5768718719482422, + "learning_rate": 4.099685439493064e-05, + "loss": 0.3359, + "num_input_tokens_seen": 3590488, + "step": 13395 + }, + { + "epoch": 3.512450851900393, + "grad_norm": 0.9695117473602295, + "learning_rate": 4.098806340212106e-05, + "loss": 0.3224, + "num_input_tokens_seen": 3591640, + "step": 13400 + }, + { + "epoch": 3.5137614678899083, + "grad_norm": 0.5537093281745911, + "learning_rate": 4.097926906303744e-05, + "loss": 0.3487, + "num_input_tokens_seen": 3592904, + "step": 13405 + }, + { + "epoch": 3.5150720838794234, + "grad_norm": 0.795621395111084, + "learning_rate": 4.0970471379520414e-05, + "loss": 0.4647, + "num_input_tokens_seen": 3594712, + "step": 13410 + }, + { + "epoch": 3.5163826998689385, + "grad_norm": 0.4190838634967804, + "learning_rate": 4.096167035341132e-05, + "loss": 0.3464, + "num_input_tokens_seen": 3595896, + "step": 13415 + }, + { + "epoch": 3.5176933158584536, + "grad_norm": 1.0607881546020508, + "learning_rate": 4.09528659865522e-05, + "loss": 0.4442, + "num_input_tokens_seen": 3597416, + "step": 13420 + }, + { + "epoch": 3.5190039318479687, + "grad_norm": 0.946493923664093, + "learning_rate": 4.094405828078579e-05, + "loss": 0.4448, + "num_input_tokens_seen": 3598664, + "step": 13425 + }, + { + "epoch": 3.5203145478374838, + "grad_norm": 0.7643872499465942, + "learning_rate": 4.093524723795554e-05, + "loss": 0.4974, + "num_input_tokens_seen": 3599784, + "step": 13430 + }, + { + "epoch": 3.5216251638269984, + "grad_norm": 0.775386393070221, + "learning_rate": 4.0926432859905574e-05, + "loss": 0.4703, + "num_input_tokens_seen": 3601336, + "step": 13435 + }, + { + "epoch": 3.522935779816514, + "grad_norm": 1.3306660652160645, + "learning_rate": 4.0917615148480726e-05, + "loss": 0.374, + "num_input_tokens_seen": 3602968, + "step": 13440 + }, + { + "epoch": 3.5242463958060286, + "grad_norm": 0.8075704574584961, + "learning_rate": 4.0908794105526537e-05, + "loss": 0.422, + "num_input_tokens_seen": 3604312, + "step": 13445 + }, + { + "epoch": 3.525557011795544, + "grad_norm": 1.1815627813339233, + "learning_rate": 4.089996973288922e-05, + "loss": 0.4349, + "num_input_tokens_seen": 3605608, + "step": 13450 + }, + { + "epoch": 3.526867627785059, + "grad_norm": 0.606239378452301, + "learning_rate": 4.089114203241572e-05, + "loss": 0.4601, + "num_input_tokens_seen": 3607448, + "step": 13455 + }, + { + "epoch": 3.528178243774574, + "grad_norm": 0.4851471781730652, + "learning_rate": 4.088231100595364e-05, + "loss": 0.4398, + "num_input_tokens_seen": 3608728, + "step": 13460 + }, + { + "epoch": 3.529488859764089, + "grad_norm": 0.7100780010223389, + "learning_rate": 4.087347665535133e-05, + "loss": 0.342, + "num_input_tokens_seen": 3609896, + "step": 13465 + }, + { + "epoch": 3.530799475753604, + "grad_norm": 0.7480551600456238, + "learning_rate": 4.0864638982457766e-05, + "loss": 0.3974, + "num_input_tokens_seen": 3611064, + "step": 13470 + }, + { + "epoch": 3.532110091743119, + "grad_norm": 1.7993896007537842, + "learning_rate": 4.085579798912268e-05, + "loss": 0.3438, + "num_input_tokens_seen": 3612280, + "step": 13475 + }, + { + "epoch": 3.5334207077326343, + "grad_norm": 0.8509131669998169, + "learning_rate": 4.084695367719647e-05, + "loss": 0.4579, + "num_input_tokens_seen": 3613464, + "step": 13480 + }, + { + "epoch": 3.5347313237221494, + "grad_norm": 1.3389204740524292, + "learning_rate": 4.083810604853023e-05, + "loss": 0.4044, + "num_input_tokens_seen": 3614936, + "step": 13485 + }, + { + "epoch": 3.5360419397116645, + "grad_norm": 1.0132758617401123, + "learning_rate": 4.082925510497577e-05, + "loss": 0.3332, + "num_input_tokens_seen": 3616232, + "step": 13490 + }, + { + "epoch": 3.5373525557011796, + "grad_norm": 0.8033169507980347, + "learning_rate": 4.0820400848385564e-05, + "loss": 0.2313, + "num_input_tokens_seen": 3617048, + "step": 13495 + }, + { + "epoch": 3.5386631716906947, + "grad_norm": 0.8934831619262695, + "learning_rate": 4.0811543280612805e-05, + "loss": 0.4266, + "num_input_tokens_seen": 3618440, + "step": 13500 + }, + { + "epoch": 3.5399737876802098, + "grad_norm": 0.758806049823761, + "learning_rate": 4.080268240351135e-05, + "loss": 0.5347, + "num_input_tokens_seen": 3619800, + "step": 13505 + }, + { + "epoch": 3.541284403669725, + "grad_norm": 0.7512359619140625, + "learning_rate": 4.079381821893576e-05, + "loss": 0.3309, + "num_input_tokens_seen": 3620984, + "step": 13510 + }, + { + "epoch": 3.54259501965924, + "grad_norm": 1.1879644393920898, + "learning_rate": 4.078495072874132e-05, + "loss": 0.3331, + "num_input_tokens_seen": 3621960, + "step": 13515 + }, + { + "epoch": 3.543905635648755, + "grad_norm": 0.7826413512229919, + "learning_rate": 4.0776079934783975e-05, + "loss": 0.2528, + "num_input_tokens_seen": 3623016, + "step": 13520 + }, + { + "epoch": 3.54521625163827, + "grad_norm": 0.5665121078491211, + "learning_rate": 4.076720583892034e-05, + "loss": 0.417, + "num_input_tokens_seen": 3624408, + "step": 13525 + }, + { + "epoch": 3.546526867627785, + "grad_norm": 2.649885416030884, + "learning_rate": 4.075832844300778e-05, + "loss": 0.3921, + "num_input_tokens_seen": 3625624, + "step": 13530 + }, + { + "epoch": 3.5478374836173003, + "grad_norm": 0.902969479560852, + "learning_rate": 4.0749447748904304e-05, + "loss": 0.3543, + "num_input_tokens_seen": 3626808, + "step": 13535 + }, + { + "epoch": 3.549148099606815, + "grad_norm": 1.9420076608657837, + "learning_rate": 4.074056375846862e-05, + "loss": 0.5306, + "num_input_tokens_seen": 3628104, + "step": 13540 + }, + { + "epoch": 3.5504587155963305, + "grad_norm": 1.6591686010360718, + "learning_rate": 4.0731676473560145e-05, + "loss": 0.3626, + "num_input_tokens_seen": 3629048, + "step": 13545 + }, + { + "epoch": 3.551769331585845, + "grad_norm": 3.8834099769592285, + "learning_rate": 4.072278589603896e-05, + "loss": 0.4778, + "num_input_tokens_seen": 3630536, + "step": 13550 + }, + { + "epoch": 3.5530799475753603, + "grad_norm": 0.6684082746505737, + "learning_rate": 4.0713892027765863e-05, + "loss": 0.2788, + "num_input_tokens_seen": 3631560, + "step": 13555 + }, + { + "epoch": 3.5543905635648754, + "grad_norm": 0.7229093909263611, + "learning_rate": 4.070499487060231e-05, + "loss": 0.3451, + "num_input_tokens_seen": 3632712, + "step": 13560 + }, + { + "epoch": 3.5557011795543905, + "grad_norm": 0.8875303268432617, + "learning_rate": 4.069609442641046e-05, + "loss": 0.4575, + "num_input_tokens_seen": 3634088, + "step": 13565 + }, + { + "epoch": 3.5570117955439056, + "grad_norm": 3.047545909881592, + "learning_rate": 4.068719069705318e-05, + "loss": 0.5027, + "num_input_tokens_seen": 3635480, + "step": 13570 + }, + { + "epoch": 3.5583224115334207, + "grad_norm": 0.6788279414176941, + "learning_rate": 4.0678283684393995e-05, + "loss": 0.519, + "num_input_tokens_seen": 3637080, + "step": 13575 + }, + { + "epoch": 3.5596330275229358, + "grad_norm": 0.8368752598762512, + "learning_rate": 4.0669373390297115e-05, + "loss": 0.4761, + "num_input_tokens_seen": 3638248, + "step": 13580 + }, + { + "epoch": 3.560943643512451, + "grad_norm": 1.0486432313919067, + "learning_rate": 4.066045981662746e-05, + "loss": 0.3503, + "num_input_tokens_seen": 3639448, + "step": 13585 + }, + { + "epoch": 3.562254259501966, + "grad_norm": 1.4260170459747314, + "learning_rate": 4.065154296525063e-05, + "loss": 0.3341, + "num_input_tokens_seen": 3640648, + "step": 13590 + }, + { + "epoch": 3.563564875491481, + "grad_norm": 0.264130562543869, + "learning_rate": 4.064262283803289e-05, + "loss": 0.3498, + "num_input_tokens_seen": 3642120, + "step": 13595 + }, + { + "epoch": 3.564875491480996, + "grad_norm": 0.8386242389678955, + "learning_rate": 4.0633699436841224e-05, + "loss": 0.4661, + "num_input_tokens_seen": 3643208, + "step": 13600 + }, + { + "epoch": 3.5661861074705112, + "grad_norm": 0.4215252697467804, + "learning_rate": 4.062477276354328e-05, + "loss": 0.3659, + "num_input_tokens_seen": 3644776, + "step": 13605 + }, + { + "epoch": 3.5674967234600263, + "grad_norm": 1.0177775621414185, + "learning_rate": 4.06158428200074e-05, + "loss": 0.3944, + "num_input_tokens_seen": 3646056, + "step": 13610 + }, + { + "epoch": 3.5688073394495414, + "grad_norm": 1.1750785112380981, + "learning_rate": 4.0606909608102595e-05, + "loss": 0.402, + "num_input_tokens_seen": 3647800, + "step": 13615 + }, + { + "epoch": 3.5701179554390565, + "grad_norm": 0.7372749447822571, + "learning_rate": 4.059797312969857e-05, + "loss": 0.4091, + "num_input_tokens_seen": 3649240, + "step": 13620 + }, + { + "epoch": 3.571428571428571, + "grad_norm": 0.6767135262489319, + "learning_rate": 4.058903338666573e-05, + "loss": 0.4816, + "num_input_tokens_seen": 3650568, + "step": 13625 + }, + { + "epoch": 3.5727391874180867, + "grad_norm": 0.35657626390457153, + "learning_rate": 4.058009038087513e-05, + "loss": 0.4822, + "num_input_tokens_seen": 3652056, + "step": 13630 + }, + { + "epoch": 3.5740498034076014, + "grad_norm": 1.1928712129592896, + "learning_rate": 4.0571144114198536e-05, + "loss": 0.3147, + "num_input_tokens_seen": 3653128, + "step": 13635 + }, + { + "epoch": 3.575360419397117, + "grad_norm": 0.8610613942146301, + "learning_rate": 4.056219458850838e-05, + "loss": 0.4863, + "num_input_tokens_seen": 3654648, + "step": 13640 + }, + { + "epoch": 3.5766710353866316, + "grad_norm": 1.3347581624984741, + "learning_rate": 4.05532418056778e-05, + "loss": 0.3349, + "num_input_tokens_seen": 3655896, + "step": 13645 + }, + { + "epoch": 3.5779816513761467, + "grad_norm": 1.6179893016815186, + "learning_rate": 4.054428576758057e-05, + "loss": 0.3856, + "num_input_tokens_seen": 3657624, + "step": 13650 + }, + { + "epoch": 3.5792922673656618, + "grad_norm": 1.4857285022735596, + "learning_rate": 4.053532647609119e-05, + "loss": 0.2934, + "num_input_tokens_seen": 3658808, + "step": 13655 + }, + { + "epoch": 3.580602883355177, + "grad_norm": 1.7253775596618652, + "learning_rate": 4.0526363933084824e-05, + "loss": 0.5789, + "num_input_tokens_seen": 3660088, + "step": 13660 + }, + { + "epoch": 3.581913499344692, + "grad_norm": 0.09487073868513107, + "learning_rate": 4.051739814043731e-05, + "loss": 0.2108, + "num_input_tokens_seen": 3661160, + "step": 13665 + }, + { + "epoch": 3.583224115334207, + "grad_norm": 1.2228764295578003, + "learning_rate": 4.0508429100025185e-05, + "loss": 0.4389, + "num_input_tokens_seen": 3662424, + "step": 13670 + }, + { + "epoch": 3.584534731323722, + "grad_norm": 1.174761414527893, + "learning_rate": 4.049945681372565e-05, + "loss": 0.3331, + "num_input_tokens_seen": 3663656, + "step": 13675 + }, + { + "epoch": 3.5858453473132372, + "grad_norm": 0.7638456225395203, + "learning_rate": 4.049048128341657e-05, + "loss": 0.2565, + "num_input_tokens_seen": 3664712, + "step": 13680 + }, + { + "epoch": 3.5871559633027523, + "grad_norm": 1.316011667251587, + "learning_rate": 4.048150251097653e-05, + "loss": 0.4023, + "num_input_tokens_seen": 3666312, + "step": 13685 + }, + { + "epoch": 3.5884665792922674, + "grad_norm": 1.3218729496002197, + "learning_rate": 4.047252049828476e-05, + "loss": 0.3888, + "num_input_tokens_seen": 3667304, + "step": 13690 + }, + { + "epoch": 3.5897771952817825, + "grad_norm": 0.8971570134162903, + "learning_rate": 4.046353524722119e-05, + "loss": 0.3845, + "num_input_tokens_seen": 3668328, + "step": 13695 + }, + { + "epoch": 3.5910878112712976, + "grad_norm": 1.0287423133850098, + "learning_rate": 4.0454546759666414e-05, + "loss": 0.5451, + "num_input_tokens_seen": 3669464, + "step": 13700 + }, + { + "epoch": 3.5923984272608127, + "grad_norm": 0.9750815033912659, + "learning_rate": 4.04455550375017e-05, + "loss": 0.3157, + "num_input_tokens_seen": 3670584, + "step": 13705 + }, + { + "epoch": 3.593709043250328, + "grad_norm": 0.7478817701339722, + "learning_rate": 4.0436560082609e-05, + "loss": 0.4267, + "num_input_tokens_seen": 3671592, + "step": 13710 + }, + { + "epoch": 3.595019659239843, + "grad_norm": 1.4973554611206055, + "learning_rate": 4.0427561896870955e-05, + "loss": 0.4472, + "num_input_tokens_seen": 3673048, + "step": 13715 + }, + { + "epoch": 3.5963302752293576, + "grad_norm": 0.46417325735092163, + "learning_rate": 4.041856048217085e-05, + "loss": 0.5738, + "num_input_tokens_seen": 3674568, + "step": 13720 + }, + { + "epoch": 3.597640891218873, + "grad_norm": 2.882753849029541, + "learning_rate": 4.040955584039269e-05, + "loss": 0.4093, + "num_input_tokens_seen": 3675608, + "step": 13725 + }, + { + "epoch": 3.5989515072083877, + "grad_norm": 0.302165687084198, + "learning_rate": 4.040054797342112e-05, + "loss": 0.3541, + "num_input_tokens_seen": 3676872, + "step": 13730 + }, + { + "epoch": 3.6002621231979033, + "grad_norm": 0.8621026873588562, + "learning_rate": 4.039153688314145e-05, + "loss": 0.3001, + "num_input_tokens_seen": 3678360, + "step": 13735 + }, + { + "epoch": 3.601572739187418, + "grad_norm": 0.8096976280212402, + "learning_rate": 4.038252257143973e-05, + "loss": 0.3875, + "num_input_tokens_seen": 3679992, + "step": 13740 + }, + { + "epoch": 3.602883355176933, + "grad_norm": 2.095799207687378, + "learning_rate": 4.03735050402026e-05, + "loss": 0.4495, + "num_input_tokens_seen": 3682120, + "step": 13745 + }, + { + "epoch": 3.604193971166448, + "grad_norm": 0.7532418370246887, + "learning_rate": 4.036448429131743e-05, + "loss": 0.3637, + "num_input_tokens_seen": 3683480, + "step": 13750 + }, + { + "epoch": 3.6055045871559632, + "grad_norm": 0.7205791473388672, + "learning_rate": 4.035546032667225e-05, + "loss": 0.3792, + "num_input_tokens_seen": 3685368, + "step": 13755 + }, + { + "epoch": 3.6068152031454783, + "grad_norm": 1.9263787269592285, + "learning_rate": 4.034643314815575e-05, + "loss": 0.5745, + "num_input_tokens_seen": 3686552, + "step": 13760 + }, + { + "epoch": 3.6081258191349934, + "grad_norm": 1.5355606079101562, + "learning_rate": 4.0337402757657314e-05, + "loss": 0.371, + "num_input_tokens_seen": 3687816, + "step": 13765 + }, + { + "epoch": 3.6094364351245085, + "grad_norm": 0.5221459269523621, + "learning_rate": 4.032836915706698e-05, + "loss": 0.3458, + "num_input_tokens_seen": 3689176, + "step": 13770 + }, + { + "epoch": 3.6107470511140236, + "grad_norm": 0.6730360388755798, + "learning_rate": 4.0319332348275465e-05, + "loss": 0.3531, + "num_input_tokens_seen": 3690232, + "step": 13775 + }, + { + "epoch": 3.6120576671035387, + "grad_norm": 1.200636386871338, + "learning_rate": 4.031029233317416e-05, + "loss": 0.331, + "num_input_tokens_seen": 3691480, + "step": 13780 + }, + { + "epoch": 3.613368283093054, + "grad_norm": 1.7685729265213013, + "learning_rate": 4.0301249113655125e-05, + "loss": 0.4524, + "num_input_tokens_seen": 3693016, + "step": 13785 + }, + { + "epoch": 3.614678899082569, + "grad_norm": 0.8761757016181946, + "learning_rate": 4.0292202691611094e-05, + "loss": 0.3734, + "num_input_tokens_seen": 3694136, + "step": 13790 + }, + { + "epoch": 3.615989515072084, + "grad_norm": 0.9425565004348755, + "learning_rate": 4.028315306893545e-05, + "loss": 0.5623, + "num_input_tokens_seen": 3695672, + "step": 13795 + }, + { + "epoch": 3.617300131061599, + "grad_norm": 0.4279540181159973, + "learning_rate": 4.0274100247522287e-05, + "loss": 0.3129, + "num_input_tokens_seen": 3697032, + "step": 13800 + }, + { + "epoch": 3.618610747051114, + "grad_norm": 1.9098494052886963, + "learning_rate": 4.026504422926632e-05, + "loss": 0.3507, + "num_input_tokens_seen": 3698088, + "step": 13805 + }, + { + "epoch": 3.6199213630406293, + "grad_norm": 0.8382517099380493, + "learning_rate": 4.025598501606299e-05, + "loss": 0.3867, + "num_input_tokens_seen": 3699624, + "step": 13810 + }, + { + "epoch": 3.621231979030144, + "grad_norm": 0.5390089154243469, + "learning_rate": 4.024692260980835e-05, + "loss": 0.3851, + "num_input_tokens_seen": 3701080, + "step": 13815 + }, + { + "epoch": 3.6225425950196595, + "grad_norm": 2.3183648586273193, + "learning_rate": 4.023785701239915e-05, + "loss": 0.4054, + "num_input_tokens_seen": 3702088, + "step": 13820 + }, + { + "epoch": 3.623853211009174, + "grad_norm": 0.9345585703849792, + "learning_rate": 4.022878822573281e-05, + "loss": 0.3359, + "num_input_tokens_seen": 3703224, + "step": 13825 + }, + { + "epoch": 3.625163826998689, + "grad_norm": 1.8674569129943848, + "learning_rate": 4.0219716251707395e-05, + "loss": 0.4456, + "num_input_tokens_seen": 3704520, + "step": 13830 + }, + { + "epoch": 3.6264744429882043, + "grad_norm": 1.1025714874267578, + "learning_rate": 4.021064109222168e-05, + "loss": 0.2641, + "num_input_tokens_seen": 3706584, + "step": 13835 + }, + { + "epoch": 3.6277850589777194, + "grad_norm": 2.0411758422851562, + "learning_rate": 4.020156274917506e-05, + "loss": 0.5227, + "num_input_tokens_seen": 3708200, + "step": 13840 + }, + { + "epoch": 3.6290956749672345, + "grad_norm": 0.577392041683197, + "learning_rate": 4.019248122446763e-05, + "loss": 0.3542, + "num_input_tokens_seen": 3709560, + "step": 13845 + }, + { + "epoch": 3.6304062909567496, + "grad_norm": 1.1455024480819702, + "learning_rate": 4.0183396520000114e-05, + "loss": 0.5539, + "num_input_tokens_seen": 3710952, + "step": 13850 + }, + { + "epoch": 3.6317169069462647, + "grad_norm": 0.7570542693138123, + "learning_rate": 4.0174308637673956e-05, + "loss": 0.5472, + "num_input_tokens_seen": 3712392, + "step": 13855 + }, + { + "epoch": 3.63302752293578, + "grad_norm": 0.9359475374221802, + "learning_rate": 4.016521757939121e-05, + "loss": 0.494, + "num_input_tokens_seen": 3713464, + "step": 13860 + }, + { + "epoch": 3.634338138925295, + "grad_norm": 1.1936602592468262, + "learning_rate": 4.015612334705464e-05, + "loss": 0.3408, + "num_input_tokens_seen": 3714968, + "step": 13865 + }, + { + "epoch": 3.63564875491481, + "grad_norm": 0.5347751975059509, + "learning_rate": 4.014702594256764e-05, + "loss": 0.5619, + "num_input_tokens_seen": 3716248, + "step": 13870 + }, + { + "epoch": 3.636959370904325, + "grad_norm": 2.5137152671813965, + "learning_rate": 4.013792536783427e-05, + "loss": 0.4333, + "num_input_tokens_seen": 3717736, + "step": 13875 + }, + { + "epoch": 3.63826998689384, + "grad_norm": 1.3422391414642334, + "learning_rate": 4.012882162475928e-05, + "loss": 0.4496, + "num_input_tokens_seen": 3718904, + "step": 13880 + }, + { + "epoch": 3.6395806028833553, + "grad_norm": 0.5019269585609436, + "learning_rate": 4.011971471524807e-05, + "loss": 0.3186, + "num_input_tokens_seen": 3720104, + "step": 13885 + }, + { + "epoch": 3.6408912188728704, + "grad_norm": 2.1926119327545166, + "learning_rate": 4.011060464120669e-05, + "loss": 0.3688, + "num_input_tokens_seen": 3721288, + "step": 13890 + }, + { + "epoch": 3.6422018348623855, + "grad_norm": 1.099453330039978, + "learning_rate": 4.010149140454188e-05, + "loss": 0.4019, + "num_input_tokens_seen": 3722856, + "step": 13895 + }, + { + "epoch": 3.6435124508519, + "grad_norm": 0.7913293838500977, + "learning_rate": 4.009237500716101e-05, + "loss": 0.3515, + "num_input_tokens_seen": 3724264, + "step": 13900 + }, + { + "epoch": 3.6448230668414157, + "grad_norm": 0.2775532603263855, + "learning_rate": 4.008325545097212e-05, + "loss": 0.1809, + "num_input_tokens_seen": 3725528, + "step": 13905 + }, + { + "epoch": 3.6461336828309303, + "grad_norm": 1.561263084411621, + "learning_rate": 4.0074132737883927e-05, + "loss": 0.4193, + "num_input_tokens_seen": 3726888, + "step": 13910 + }, + { + "epoch": 3.647444298820446, + "grad_norm": 3.199291706085205, + "learning_rate": 4.00650068698058e-05, + "loss": 0.5271, + "num_input_tokens_seen": 3728472, + "step": 13915 + }, + { + "epoch": 3.6487549148099605, + "grad_norm": 0.9362967610359192, + "learning_rate": 4.005587784864776e-05, + "loss": 0.4862, + "num_input_tokens_seen": 3729704, + "step": 13920 + }, + { + "epoch": 3.6500655307994756, + "grad_norm": 1.0699045658111572, + "learning_rate": 4.0046745676320504e-05, + "loss": 0.3618, + "num_input_tokens_seen": 3731144, + "step": 13925 + }, + { + "epoch": 3.6513761467889907, + "grad_norm": 1.4877921342849731, + "learning_rate": 4.003761035473538e-05, + "loss": 0.4771, + "num_input_tokens_seen": 3732456, + "step": 13930 + }, + { + "epoch": 3.652686762778506, + "grad_norm": 0.7341650724411011, + "learning_rate": 4.002847188580439e-05, + "loss": 0.4796, + "num_input_tokens_seen": 3734312, + "step": 13935 + }, + { + "epoch": 3.653997378768021, + "grad_norm": 1.0897581577301025, + "learning_rate": 4.0019330271440206e-05, + "loss": 0.4003, + "num_input_tokens_seen": 3735944, + "step": 13940 + }, + { + "epoch": 3.655307994757536, + "grad_norm": 0.5856874585151672, + "learning_rate": 4.001018551355614e-05, + "loss": 0.3611, + "num_input_tokens_seen": 3737304, + "step": 13945 + }, + { + "epoch": 3.656618610747051, + "grad_norm": 0.43117693066596985, + "learning_rate": 4.0001037614066184e-05, + "loss": 0.3251, + "num_input_tokens_seen": 3738696, + "step": 13950 + }, + { + "epoch": 3.657929226736566, + "grad_norm": 0.9334466457366943, + "learning_rate": 3.9991886574884975e-05, + "loss": 0.4176, + "num_input_tokens_seen": 3739752, + "step": 13955 + }, + { + "epoch": 3.6592398427260813, + "grad_norm": 1.0173816680908203, + "learning_rate": 3.998273239792781e-05, + "loss": 0.4406, + "num_input_tokens_seen": 3741176, + "step": 13960 + }, + { + "epoch": 3.6605504587155964, + "grad_norm": 1.399106740951538, + "learning_rate": 3.997357508511064e-05, + "loss": 0.39, + "num_input_tokens_seen": 3742472, + "step": 13965 + }, + { + "epoch": 3.6618610747051115, + "grad_norm": 0.7580487132072449, + "learning_rate": 3.996441463835008e-05, + "loss": 0.4158, + "num_input_tokens_seen": 3743592, + "step": 13970 + }, + { + "epoch": 3.6631716906946266, + "grad_norm": 0.7259737849235535, + "learning_rate": 3.995525105956339e-05, + "loss": 0.3284, + "num_input_tokens_seen": 3744888, + "step": 13975 + }, + { + "epoch": 3.6644823066841417, + "grad_norm": 1.0641179084777832, + "learning_rate": 3.9946084350668506e-05, + "loss": 0.4235, + "num_input_tokens_seen": 3745928, + "step": 13980 + }, + { + "epoch": 3.6657929226736568, + "grad_norm": 1.3066200017929077, + "learning_rate": 3.993691451358398e-05, + "loss": 0.4785, + "num_input_tokens_seen": 3746936, + "step": 13985 + }, + { + "epoch": 3.667103538663172, + "grad_norm": 0.5849630236625671, + "learning_rate": 3.992774155022906e-05, + "loss": 0.3556, + "num_input_tokens_seen": 3748600, + "step": 13990 + }, + { + "epoch": 3.6684141546526865, + "grad_norm": 1.3685892820358276, + "learning_rate": 3.991856546252362e-05, + "loss": 0.4946, + "num_input_tokens_seen": 3750104, + "step": 13995 + }, + { + "epoch": 3.669724770642202, + "grad_norm": 0.8797978758811951, + "learning_rate": 3.9909386252388215e-05, + "loss": 0.3474, + "num_input_tokens_seen": 3751752, + "step": 14000 + }, + { + "epoch": 3.6710353866317167, + "grad_norm": 1.5342410802841187, + "learning_rate": 3.990020392174402e-05, + "loss": 0.317, + "num_input_tokens_seen": 3753224, + "step": 14005 + }, + { + "epoch": 3.6723460026212322, + "grad_norm": 1.6586743593215942, + "learning_rate": 3.9891018472512895e-05, + "loss": 0.4909, + "num_input_tokens_seen": 3754280, + "step": 14010 + }, + { + "epoch": 3.673656618610747, + "grad_norm": 0.5664445757865906, + "learning_rate": 3.988182990661734e-05, + "loss": 0.3633, + "num_input_tokens_seen": 3755640, + "step": 14015 + }, + { + "epoch": 3.674967234600262, + "grad_norm": 0.5656618475914001, + "learning_rate": 3.987263822598049e-05, + "loss": 0.3745, + "num_input_tokens_seen": 3757064, + "step": 14020 + }, + { + "epoch": 3.676277850589777, + "grad_norm": 0.6694218516349792, + "learning_rate": 3.986344343252615e-05, + "loss": 0.3356, + "num_input_tokens_seen": 3758296, + "step": 14025 + }, + { + "epoch": 3.677588466579292, + "grad_norm": 0.7234585881233215, + "learning_rate": 3.9854245528178804e-05, + "loss": 0.3327, + "num_input_tokens_seen": 3759496, + "step": 14030 + }, + { + "epoch": 3.6788990825688073, + "grad_norm": 2.710676670074463, + "learning_rate": 3.984504451486352e-05, + "loss": 0.6242, + "num_input_tokens_seen": 3760712, + "step": 14035 + }, + { + "epoch": 3.6802096985583224, + "grad_norm": 0.9722538590431213, + "learning_rate": 3.983584039450607e-05, + "loss": 0.4212, + "num_input_tokens_seen": 3761800, + "step": 14040 + }, + { + "epoch": 3.6815203145478375, + "grad_norm": 1.171012043952942, + "learning_rate": 3.982663316903286e-05, + "loss": 0.3058, + "num_input_tokens_seen": 3763368, + "step": 14045 + }, + { + "epoch": 3.6828309305373526, + "grad_norm": 1.029745101928711, + "learning_rate": 3.981742284037095e-05, + "loss": 0.4311, + "num_input_tokens_seen": 3764632, + "step": 14050 + }, + { + "epoch": 3.6841415465268676, + "grad_norm": 0.8402113318443298, + "learning_rate": 3.980820941044803e-05, + "loss": 0.4396, + "num_input_tokens_seen": 3765896, + "step": 14055 + }, + { + "epoch": 3.6854521625163827, + "grad_norm": 1.1324161291122437, + "learning_rate": 3.979899288119248e-05, + "loss": 0.3214, + "num_input_tokens_seen": 3767128, + "step": 14060 + }, + { + "epoch": 3.686762778505898, + "grad_norm": 1.0217576026916504, + "learning_rate": 3.978977325453329e-05, + "loss": 0.3134, + "num_input_tokens_seen": 3768136, + "step": 14065 + }, + { + "epoch": 3.688073394495413, + "grad_norm": 1.2992274761199951, + "learning_rate": 3.9780550532400106e-05, + "loss": 0.3557, + "num_input_tokens_seen": 3769048, + "step": 14070 + }, + { + "epoch": 3.689384010484928, + "grad_norm": 0.5174189209938049, + "learning_rate": 3.9771324716723246e-05, + "loss": 0.385, + "num_input_tokens_seen": 3770312, + "step": 14075 + }, + { + "epoch": 3.690694626474443, + "grad_norm": 2.268697500228882, + "learning_rate": 3.976209580943363e-05, + "loss": 0.4749, + "num_input_tokens_seen": 3771304, + "step": 14080 + }, + { + "epoch": 3.6920052424639582, + "grad_norm": 0.537927508354187, + "learning_rate": 3.975286381246288e-05, + "loss": 0.4882, + "num_input_tokens_seen": 3772616, + "step": 14085 + }, + { + "epoch": 3.693315858453473, + "grad_norm": 0.7952347993850708, + "learning_rate": 3.9743628727743224e-05, + "loss": 0.4535, + "num_input_tokens_seen": 3773704, + "step": 14090 + }, + { + "epoch": 3.6946264744429884, + "grad_norm": 1.5137591361999512, + "learning_rate": 3.973439055720755e-05, + "loss": 0.3495, + "num_input_tokens_seen": 3774728, + "step": 14095 + }, + { + "epoch": 3.695937090432503, + "grad_norm": 1.5788933038711548, + "learning_rate": 3.972514930278938e-05, + "loss": 0.8127, + "num_input_tokens_seen": 3776440, + "step": 14100 + }, + { + "epoch": 3.6972477064220186, + "grad_norm": 0.678059458732605, + "learning_rate": 3.971590496642291e-05, + "loss": 0.3512, + "num_input_tokens_seen": 3777768, + "step": 14105 + }, + { + "epoch": 3.6985583224115333, + "grad_norm": 1.1478168964385986, + "learning_rate": 3.970665755004296e-05, + "loss": 0.3671, + "num_input_tokens_seen": 3778824, + "step": 14110 + }, + { + "epoch": 3.6998689384010484, + "grad_norm": 0.4526546001434326, + "learning_rate": 3.9697407055585e-05, + "loss": 0.3141, + "num_input_tokens_seen": 3780344, + "step": 14115 + }, + { + "epoch": 3.7011795543905635, + "grad_norm": 1.7121022939682007, + "learning_rate": 3.9688153484985135e-05, + "loss": 0.4146, + "num_input_tokens_seen": 3781384, + "step": 14120 + }, + { + "epoch": 3.7024901703800785, + "grad_norm": 1.1635007858276367, + "learning_rate": 3.9678896840180125e-05, + "loss": 1.8359, + "num_input_tokens_seen": 3782488, + "step": 14125 + }, + { + "epoch": 3.7038007863695936, + "grad_norm": 0.5366883277893066, + "learning_rate": 3.9669637123107374e-05, + "loss": 0.4738, + "num_input_tokens_seen": 3784408, + "step": 14130 + }, + { + "epoch": 3.7051114023591087, + "grad_norm": 0.9486504197120667, + "learning_rate": 3.9660374335704906e-05, + "loss": 0.4437, + "num_input_tokens_seen": 3785608, + "step": 14135 + }, + { + "epoch": 3.706422018348624, + "grad_norm": 0.8315372467041016, + "learning_rate": 3.965110847991143e-05, + "loss": 0.4589, + "num_input_tokens_seen": 3786712, + "step": 14140 + }, + { + "epoch": 3.707732634338139, + "grad_norm": 0.7500312924385071, + "learning_rate": 3.964183955766627e-05, + "loss": 0.4379, + "num_input_tokens_seen": 3787928, + "step": 14145 + }, + { + "epoch": 3.709043250327654, + "grad_norm": 0.6099868416786194, + "learning_rate": 3.963256757090938e-05, + "loss": 0.4182, + "num_input_tokens_seen": 3789480, + "step": 14150 + }, + { + "epoch": 3.710353866317169, + "grad_norm": 0.8987030386924744, + "learning_rate": 3.962329252158139e-05, + "loss": 0.218, + "num_input_tokens_seen": 3790696, + "step": 14155 + }, + { + "epoch": 3.711664482306684, + "grad_norm": 1.0544732809066772, + "learning_rate": 3.961401441162354e-05, + "loss": 0.3692, + "num_input_tokens_seen": 3791832, + "step": 14160 + }, + { + "epoch": 3.7129750982961993, + "grad_norm": 0.9504876732826233, + "learning_rate": 3.960473324297772e-05, + "loss": 0.4894, + "num_input_tokens_seen": 3792888, + "step": 14165 + }, + { + "epoch": 3.7142857142857144, + "grad_norm": 0.7998362183570862, + "learning_rate": 3.959544901758646e-05, + "loss": 0.4315, + "num_input_tokens_seen": 3794792, + "step": 14170 + }, + { + "epoch": 3.7155963302752295, + "grad_norm": 1.802215337753296, + "learning_rate": 3.958616173739295e-05, + "loss": 0.6587, + "num_input_tokens_seen": 3796712, + "step": 14175 + }, + { + "epoch": 3.7169069462647446, + "grad_norm": 0.9196909070014954, + "learning_rate": 3.9576871404340994e-05, + "loss": 0.4677, + "num_input_tokens_seen": 3797896, + "step": 14180 + }, + { + "epoch": 3.7182175622542593, + "grad_norm": 0.717089056968689, + "learning_rate": 3.9567578020375036e-05, + "loss": 0.6573, + "num_input_tokens_seen": 3798808, + "step": 14185 + }, + { + "epoch": 3.719528178243775, + "grad_norm": 1.0268304347991943, + "learning_rate": 3.955828158744017e-05, + "loss": 0.3764, + "num_input_tokens_seen": 3800264, + "step": 14190 + }, + { + "epoch": 3.7208387942332894, + "grad_norm": 0.5743920207023621, + "learning_rate": 3.954898210748211e-05, + "loss": 0.3211, + "num_input_tokens_seen": 3802008, + "step": 14195 + }, + { + "epoch": 3.722149410222805, + "grad_norm": 1.1032607555389404, + "learning_rate": 3.953967958244725e-05, + "loss": 0.3357, + "num_input_tokens_seen": 3803256, + "step": 14200 + }, + { + "epoch": 3.7234600262123196, + "grad_norm": 1.1762653589248657, + "learning_rate": 3.9530374014282574e-05, + "loss": 0.6069, + "num_input_tokens_seen": 3804504, + "step": 14205 + }, + { + "epoch": 3.7247706422018347, + "grad_norm": 1.368952751159668, + "learning_rate": 3.952106540493572e-05, + "loss": 0.4425, + "num_input_tokens_seen": 3806424, + "step": 14210 + }, + { + "epoch": 3.72608125819135, + "grad_norm": 1.0229051113128662, + "learning_rate": 3.951175375635497e-05, + "loss": 0.2906, + "num_input_tokens_seen": 3807752, + "step": 14215 + }, + { + "epoch": 3.727391874180865, + "grad_norm": 0.7201188802719116, + "learning_rate": 3.9502439070489224e-05, + "loss": 0.3971, + "num_input_tokens_seen": 3808856, + "step": 14220 + }, + { + "epoch": 3.72870249017038, + "grad_norm": 0.6981839537620544, + "learning_rate": 3.949312134928805e-05, + "loss": 0.5236, + "num_input_tokens_seen": 3810040, + "step": 14225 + }, + { + "epoch": 3.730013106159895, + "grad_norm": 0.7544784545898438, + "learning_rate": 3.948380059470161e-05, + "loss": 0.372, + "num_input_tokens_seen": 3811224, + "step": 14230 + }, + { + "epoch": 3.73132372214941, + "grad_norm": 1.4562045335769653, + "learning_rate": 3.947447680868074e-05, + "loss": 0.6475, + "num_input_tokens_seen": 3812424, + "step": 14235 + }, + { + "epoch": 3.7326343381389253, + "grad_norm": 1.212599277496338, + "learning_rate": 3.9465149993176885e-05, + "loss": 0.4377, + "num_input_tokens_seen": 3813576, + "step": 14240 + }, + { + "epoch": 3.7339449541284404, + "grad_norm": 0.8238811492919922, + "learning_rate": 3.945582015014212e-05, + "loss": 0.3661, + "num_input_tokens_seen": 3814536, + "step": 14245 + }, + { + "epoch": 3.7352555701179555, + "grad_norm": 0.9165405035018921, + "learning_rate": 3.944648728152919e-05, + "loss": 0.3776, + "num_input_tokens_seen": 3816504, + "step": 14250 + }, + { + "epoch": 3.7365661861074706, + "grad_norm": 0.8222938776016235, + "learning_rate": 3.943715138929142e-05, + "loss": 0.4027, + "num_input_tokens_seen": 3817640, + "step": 14255 + }, + { + "epoch": 3.7378768020969857, + "grad_norm": 2.053452491760254, + "learning_rate": 3.942781247538282e-05, + "loss": 0.4671, + "num_input_tokens_seen": 3818696, + "step": 14260 + }, + { + "epoch": 3.739187418086501, + "grad_norm": 2.4997434616088867, + "learning_rate": 3.941847054175799e-05, + "loss": 0.2391, + "num_input_tokens_seen": 3819736, + "step": 14265 + }, + { + "epoch": 3.740498034076016, + "grad_norm": 0.7941008806228638, + "learning_rate": 3.940912559037219e-05, + "loss": 0.5816, + "num_input_tokens_seen": 3821256, + "step": 14270 + }, + { + "epoch": 3.741808650065531, + "grad_norm": 1.1027987003326416, + "learning_rate": 3.939977762318131e-05, + "loss": 0.4451, + "num_input_tokens_seen": 3822472, + "step": 14275 + }, + { + "epoch": 3.7431192660550456, + "grad_norm": 2.3250975608825684, + "learning_rate": 3.939042664214184e-05, + "loss": 0.4163, + "num_input_tokens_seen": 3823400, + "step": 14280 + }, + { + "epoch": 3.744429882044561, + "grad_norm": 1.4702354669570923, + "learning_rate": 3.938107264921095e-05, + "loss": 0.5399, + "num_input_tokens_seen": 3824472, + "step": 14285 + }, + { + "epoch": 3.745740498034076, + "grad_norm": 0.907719075679779, + "learning_rate": 3.9371715646346396e-05, + "loss": 0.2439, + "num_input_tokens_seen": 3826408, + "step": 14290 + }, + { + "epoch": 3.747051114023591, + "grad_norm": 0.7121026515960693, + "learning_rate": 3.936235563550659e-05, + "loss": 0.4557, + "num_input_tokens_seen": 3828456, + "step": 14295 + }, + { + "epoch": 3.748361730013106, + "grad_norm": 0.8924551010131836, + "learning_rate": 3.935299261865057e-05, + "loss": 0.3607, + "num_input_tokens_seen": 3829640, + "step": 14300 + }, + { + "epoch": 3.749672346002621, + "grad_norm": 0.6573741436004639, + "learning_rate": 3.934362659773799e-05, + "loss": 0.3436, + "num_input_tokens_seen": 3830680, + "step": 14305 + }, + { + "epoch": 3.750982961992136, + "grad_norm": 2.668114185333252, + "learning_rate": 3.933425757472915e-05, + "loss": 0.4396, + "num_input_tokens_seen": 3831768, + "step": 14310 + }, + { + "epoch": 3.7522935779816513, + "grad_norm": 0.9926148653030396, + "learning_rate": 3.932488555158497e-05, + "loss": 0.493, + "num_input_tokens_seen": 3833272, + "step": 14315 + }, + { + "epoch": 3.7536041939711664, + "grad_norm": 0.9685218930244446, + "learning_rate": 3.931551053026699e-05, + "loss": 0.2754, + "num_input_tokens_seen": 3834408, + "step": 14320 + }, + { + "epoch": 3.7549148099606815, + "grad_norm": 0.910666286945343, + "learning_rate": 3.9306132512737395e-05, + "loss": 0.3466, + "num_input_tokens_seen": 3835432, + "step": 14325 + }, + { + "epoch": 3.7562254259501966, + "grad_norm": 3.119248628616333, + "learning_rate": 3.929675150095898e-05, + "loss": 0.3127, + "num_input_tokens_seen": 3836808, + "step": 14330 + }, + { + "epoch": 3.7575360419397117, + "grad_norm": 0.6545151472091675, + "learning_rate": 3.928736749689519e-05, + "loss": 0.3816, + "num_input_tokens_seen": 3837976, + "step": 14335 + }, + { + "epoch": 3.758846657929227, + "grad_norm": 0.681968092918396, + "learning_rate": 3.927798050251006e-05, + "loss": 0.328, + "num_input_tokens_seen": 3839288, + "step": 14340 + }, + { + "epoch": 3.760157273918742, + "grad_norm": 0.9967012405395508, + "learning_rate": 3.926859051976828e-05, + "loss": 0.4451, + "num_input_tokens_seen": 3840600, + "step": 14345 + }, + { + "epoch": 3.761467889908257, + "grad_norm": 0.8848848342895508, + "learning_rate": 3.9259197550635164e-05, + "loss": 0.4105, + "num_input_tokens_seen": 3841704, + "step": 14350 + }, + { + "epoch": 3.762778505897772, + "grad_norm": 3.1289663314819336, + "learning_rate": 3.924980159707664e-05, + "loss": 0.5711, + "num_input_tokens_seen": 3842776, + "step": 14355 + }, + { + "epoch": 3.764089121887287, + "grad_norm": 1.0464625358581543, + "learning_rate": 3.924040266105926e-05, + "loss": 0.2611, + "num_input_tokens_seen": 3844072, + "step": 14360 + }, + { + "epoch": 3.765399737876802, + "grad_norm": 0.9028819799423218, + "learning_rate": 3.9231000744550205e-05, + "loss": 0.426, + "num_input_tokens_seen": 3845208, + "step": 14365 + }, + { + "epoch": 3.7667103538663174, + "grad_norm": 0.4151974320411682, + "learning_rate": 3.922159584951729e-05, + "loss": 0.4567, + "num_input_tokens_seen": 3846584, + "step": 14370 + }, + { + "epoch": 3.768020969855832, + "grad_norm": 0.6580379605293274, + "learning_rate": 3.921218797792893e-05, + "loss": 0.4338, + "num_input_tokens_seen": 3848040, + "step": 14375 + }, + { + "epoch": 3.7693315858453476, + "grad_norm": 1.4543286561965942, + "learning_rate": 3.9202777131754187e-05, + "loss": 0.4665, + "num_input_tokens_seen": 3849224, + "step": 14380 + }, + { + "epoch": 3.770642201834862, + "grad_norm": 0.6627604365348816, + "learning_rate": 3.9193363312962725e-05, + "loss": 0.3963, + "num_input_tokens_seen": 3850424, + "step": 14385 + }, + { + "epoch": 3.7719528178243773, + "grad_norm": 0.8060852289199829, + "learning_rate": 3.9183946523524856e-05, + "loss": 0.3696, + "num_input_tokens_seen": 3852008, + "step": 14390 + }, + { + "epoch": 3.7732634338138924, + "grad_norm": 0.45047372579574585, + "learning_rate": 3.917452676541148e-05, + "loss": 0.4673, + "num_input_tokens_seen": 3853160, + "step": 14395 + }, + { + "epoch": 3.7745740498034075, + "grad_norm": 0.9069773554801941, + "learning_rate": 3.916510404059415e-05, + "loss": 0.4125, + "num_input_tokens_seen": 3854568, + "step": 14400 + }, + { + "epoch": 3.7758846657929226, + "grad_norm": 1.0102933645248413, + "learning_rate": 3.9155678351045014e-05, + "loss": 0.4799, + "num_input_tokens_seen": 3856280, + "step": 14405 + }, + { + "epoch": 3.7771952817824377, + "grad_norm": 1.2067102193832397, + "learning_rate": 3.914624969873686e-05, + "loss": 0.3305, + "num_input_tokens_seen": 3857672, + "step": 14410 + }, + { + "epoch": 3.778505897771953, + "grad_norm": 4.191128253936768, + "learning_rate": 3.913681808564309e-05, + "loss": 0.4973, + "num_input_tokens_seen": 3858904, + "step": 14415 + }, + { + "epoch": 3.779816513761468, + "grad_norm": 0.8517662286758423, + "learning_rate": 3.912738351373772e-05, + "loss": 0.4636, + "num_input_tokens_seen": 3860344, + "step": 14420 + }, + { + "epoch": 3.781127129750983, + "grad_norm": 1.907017707824707, + "learning_rate": 3.911794598499539e-05, + "loss": 0.4521, + "num_input_tokens_seen": 3861368, + "step": 14425 + }, + { + "epoch": 3.782437745740498, + "grad_norm": 1.076869010925293, + "learning_rate": 3.9108505501391355e-05, + "loss": 0.4407, + "num_input_tokens_seen": 3862504, + "step": 14430 + }, + { + "epoch": 3.783748361730013, + "grad_norm": 0.7843132019042969, + "learning_rate": 3.9099062064901497e-05, + "loss": 0.361, + "num_input_tokens_seen": 3863752, + "step": 14435 + }, + { + "epoch": 3.7850589777195283, + "grad_norm": 0.8875554203987122, + "learning_rate": 3.908961567750231e-05, + "loss": 0.3239, + "num_input_tokens_seen": 3865176, + "step": 14440 + }, + { + "epoch": 3.7863695937090434, + "grad_norm": 1.2097089290618896, + "learning_rate": 3.908016634117092e-05, + "loss": 0.4267, + "num_input_tokens_seen": 3866392, + "step": 14445 + }, + { + "epoch": 3.7876802096985585, + "grad_norm": 0.6239581108093262, + "learning_rate": 3.907071405788503e-05, + "loss": 0.4309, + "num_input_tokens_seen": 3867416, + "step": 14450 + }, + { + "epoch": 3.7889908256880735, + "grad_norm": 2.352497100830078, + "learning_rate": 3.9061258829622995e-05, + "loss": 0.4843, + "num_input_tokens_seen": 3868712, + "step": 14455 + }, + { + "epoch": 3.790301441677588, + "grad_norm": 0.5538812279701233, + "learning_rate": 3.905180065836379e-05, + "loss": 0.2768, + "num_input_tokens_seen": 3870152, + "step": 14460 + }, + { + "epoch": 3.7916120576671037, + "grad_norm": 2.3460700511932373, + "learning_rate": 3.904233954608699e-05, + "loss": 0.4391, + "num_input_tokens_seen": 3871480, + "step": 14465 + }, + { + "epoch": 3.7929226736566184, + "grad_norm": 1.2043161392211914, + "learning_rate": 3.903287549477279e-05, + "loss": 0.291, + "num_input_tokens_seen": 3872792, + "step": 14470 + }, + { + "epoch": 3.794233289646134, + "grad_norm": 1.0918279886245728, + "learning_rate": 3.9023408506401987e-05, + "loss": 0.4979, + "num_input_tokens_seen": 3873944, + "step": 14475 + }, + { + "epoch": 3.7955439056356486, + "grad_norm": 1.0488899946212769, + "learning_rate": 3.901393858295602e-05, + "loss": 0.5417, + "num_input_tokens_seen": 3876200, + "step": 14480 + }, + { + "epoch": 3.7968545216251637, + "grad_norm": 0.9977442622184753, + "learning_rate": 3.900446572641692e-05, + "loss": 0.4321, + "num_input_tokens_seen": 3877928, + "step": 14485 + }, + { + "epoch": 3.7981651376146788, + "grad_norm": 0.7472956776618958, + "learning_rate": 3.899498993876733e-05, + "loss": 0.4813, + "num_input_tokens_seen": 3878904, + "step": 14490 + }, + { + "epoch": 3.799475753604194, + "grad_norm": 0.6600097417831421, + "learning_rate": 3.898551122199054e-05, + "loss": 0.3376, + "num_input_tokens_seen": 3880136, + "step": 14495 + }, + { + "epoch": 3.800786369593709, + "grad_norm": 0.5920346975326538, + "learning_rate": 3.8976029578070415e-05, + "loss": 0.4757, + "num_input_tokens_seen": 3881272, + "step": 14500 + }, + { + "epoch": 3.802096985583224, + "grad_norm": 0.6639214754104614, + "learning_rate": 3.896654500899145e-05, + "loss": 0.4246, + "num_input_tokens_seen": 3882504, + "step": 14505 + }, + { + "epoch": 3.803407601572739, + "grad_norm": 0.9151773452758789, + "learning_rate": 3.895705751673874e-05, + "loss": 0.4896, + "num_input_tokens_seen": 3885368, + "step": 14510 + }, + { + "epoch": 3.8047182175622543, + "grad_norm": 0.7773666977882385, + "learning_rate": 3.894756710329801e-05, + "loss": 0.5492, + "num_input_tokens_seen": 3886904, + "step": 14515 + }, + { + "epoch": 3.8060288335517694, + "grad_norm": 1.153709053993225, + "learning_rate": 3.893807377065559e-05, + "loss": 0.439, + "num_input_tokens_seen": 3887976, + "step": 14520 + }, + { + "epoch": 3.8073394495412844, + "grad_norm": 0.7205227017402649, + "learning_rate": 3.8928577520798407e-05, + "loss": 0.3577, + "num_input_tokens_seen": 3889064, + "step": 14525 + }, + { + "epoch": 3.8086500655307995, + "grad_norm": 0.657290518283844, + "learning_rate": 3.8919078355714025e-05, + "loss": 0.2164, + "num_input_tokens_seen": 3890344, + "step": 14530 + }, + { + "epoch": 3.8099606815203146, + "grad_norm": 0.8829153776168823, + "learning_rate": 3.890957627739058e-05, + "loss": 0.4566, + "num_input_tokens_seen": 3892056, + "step": 14535 + }, + { + "epoch": 3.8112712975098297, + "grad_norm": 0.6634449362754822, + "learning_rate": 3.890007128781686e-05, + "loss": 0.2756, + "num_input_tokens_seen": 3893304, + "step": 14540 + }, + { + "epoch": 3.812581913499345, + "grad_norm": 0.8623570799827576, + "learning_rate": 3.889056338898224e-05, + "loss": 0.2531, + "num_input_tokens_seen": 3894872, + "step": 14545 + }, + { + "epoch": 3.81389252948886, + "grad_norm": 1.118747353553772, + "learning_rate": 3.8881052582876695e-05, + "loss": 0.5913, + "num_input_tokens_seen": 3896504, + "step": 14550 + }, + { + "epoch": 3.8152031454783746, + "grad_norm": 1.4036259651184082, + "learning_rate": 3.887153887149084e-05, + "loss": 0.3779, + "num_input_tokens_seen": 3897768, + "step": 14555 + }, + { + "epoch": 3.81651376146789, + "grad_norm": 0.8053616881370544, + "learning_rate": 3.8862022256815865e-05, + "loss": 0.3325, + "num_input_tokens_seen": 3898872, + "step": 14560 + }, + { + "epoch": 3.8178243774574048, + "grad_norm": 1.1441715955734253, + "learning_rate": 3.885250274084358e-05, + "loss": 0.3993, + "num_input_tokens_seen": 3900008, + "step": 14565 + }, + { + "epoch": 3.8191349934469203, + "grad_norm": 0.9120131134986877, + "learning_rate": 3.88429803255664e-05, + "loss": 0.3696, + "num_input_tokens_seen": 3901064, + "step": 14570 + }, + { + "epoch": 3.820445609436435, + "grad_norm": 1.3276865482330322, + "learning_rate": 3.883345501297737e-05, + "loss": 0.3467, + "num_input_tokens_seen": 3902568, + "step": 14575 + }, + { + "epoch": 3.82175622542595, + "grad_norm": 1.160256266593933, + "learning_rate": 3.88239268050701e-05, + "loss": 0.6545, + "num_input_tokens_seen": 3904088, + "step": 14580 + }, + { + "epoch": 3.823066841415465, + "grad_norm": 1.3077086210250854, + "learning_rate": 3.881439570383884e-05, + "loss": 0.5891, + "num_input_tokens_seen": 3905464, + "step": 14585 + }, + { + "epoch": 3.8243774574049803, + "grad_norm": 0.4523071348667145, + "learning_rate": 3.8804861711278426e-05, + "loss": 0.3781, + "num_input_tokens_seen": 3906952, + "step": 14590 + }, + { + "epoch": 3.8256880733944953, + "grad_norm": 1.426105260848999, + "learning_rate": 3.879532482938431e-05, + "loss": 0.4621, + "num_input_tokens_seen": 3908056, + "step": 14595 + }, + { + "epoch": 3.8269986893840104, + "grad_norm": 0.6824497580528259, + "learning_rate": 3.878578506015254e-05, + "loss": 0.3516, + "num_input_tokens_seen": 3909256, + "step": 14600 + }, + { + "epoch": 3.8283093053735255, + "grad_norm": 1.036712884902954, + "learning_rate": 3.877624240557978e-05, + "loss": 0.5412, + "num_input_tokens_seen": 3910424, + "step": 14605 + }, + { + "epoch": 3.8296199213630406, + "grad_norm": 0.9788817763328552, + "learning_rate": 3.876669686766329e-05, + "loss": 0.5176, + "num_input_tokens_seen": 3912536, + "step": 14610 + }, + { + "epoch": 3.8309305373525557, + "grad_norm": 1.6052368879318237, + "learning_rate": 3.875714844840093e-05, + "loss": 0.4006, + "num_input_tokens_seen": 3913688, + "step": 14615 + }, + { + "epoch": 3.832241153342071, + "grad_norm": 1.3815107345581055, + "learning_rate": 3.874759714979117e-05, + "loss": 0.3528, + "num_input_tokens_seen": 3914616, + "step": 14620 + }, + { + "epoch": 3.833551769331586, + "grad_norm": 3.593656539916992, + "learning_rate": 3.873804297383308e-05, + "loss": 0.4379, + "num_input_tokens_seen": 3916120, + "step": 14625 + }, + { + "epoch": 3.834862385321101, + "grad_norm": 1.7962939739227295, + "learning_rate": 3.872848592252634e-05, + "loss": 0.3867, + "num_input_tokens_seen": 3919400, + "step": 14630 + }, + { + "epoch": 3.836173001310616, + "grad_norm": 1.2058604955673218, + "learning_rate": 3.871892599787121e-05, + "loss": 0.4238, + "num_input_tokens_seen": 3920408, + "step": 14635 + }, + { + "epoch": 3.837483617300131, + "grad_norm": 1.804356575012207, + "learning_rate": 3.8709363201868576e-05, + "loss": 0.4586, + "num_input_tokens_seen": 3921640, + "step": 14640 + }, + { + "epoch": 3.8387942332896463, + "grad_norm": 1.4031935930252075, + "learning_rate": 3.869979753651991e-05, + "loss": 0.459, + "num_input_tokens_seen": 3922744, + "step": 14645 + }, + { + "epoch": 3.840104849279161, + "grad_norm": 0.785376250743866, + "learning_rate": 3.869022900382729e-05, + "loss": 0.2332, + "num_input_tokens_seen": 3924152, + "step": 14650 + }, + { + "epoch": 3.8414154652686765, + "grad_norm": 0.9542138576507568, + "learning_rate": 3.868065760579339e-05, + "loss": 0.3169, + "num_input_tokens_seen": 3925848, + "step": 14655 + }, + { + "epoch": 3.842726081258191, + "grad_norm": 0.7205769419670105, + "learning_rate": 3.86710833444215e-05, + "loss": 0.3904, + "num_input_tokens_seen": 3927400, + "step": 14660 + }, + { + "epoch": 3.8440366972477067, + "grad_norm": 1.320682168006897, + "learning_rate": 3.866150622171549e-05, + "loss": 0.4672, + "num_input_tokens_seen": 3928536, + "step": 14665 + }, + { + "epoch": 3.8453473132372213, + "grad_norm": 7.2457756996154785, + "learning_rate": 3.8651926239679824e-05, + "loss": 0.5085, + "num_input_tokens_seen": 3929544, + "step": 14670 + }, + { + "epoch": 3.8466579292267364, + "grad_norm": 0.6548526287078857, + "learning_rate": 3.8642343400319594e-05, + "loss": 0.4574, + "num_input_tokens_seen": 3931096, + "step": 14675 + }, + { + "epoch": 3.8479685452162515, + "grad_norm": 1.0340877771377563, + "learning_rate": 3.863275770564046e-05, + "loss": 0.3596, + "num_input_tokens_seen": 3932568, + "step": 14680 + }, + { + "epoch": 3.8492791612057666, + "grad_norm": 0.5180417895317078, + "learning_rate": 3.862316915764869e-05, + "loss": 0.2575, + "num_input_tokens_seen": 3933752, + "step": 14685 + }, + { + "epoch": 3.8505897771952817, + "grad_norm": 0.6045240163803101, + "learning_rate": 3.861357775835115e-05, + "loss": 0.4179, + "num_input_tokens_seen": 3935000, + "step": 14690 + }, + { + "epoch": 3.851900393184797, + "grad_norm": 1.2899476289749146, + "learning_rate": 3.8603983509755316e-05, + "loss": 0.3268, + "num_input_tokens_seen": 3936280, + "step": 14695 + }, + { + "epoch": 3.853211009174312, + "grad_norm": 0.6104092001914978, + "learning_rate": 3.8594386413869235e-05, + "loss": 0.3301, + "num_input_tokens_seen": 3937912, + "step": 14700 + }, + { + "epoch": 3.854521625163827, + "grad_norm": 0.9123896956443787, + "learning_rate": 3.8584786472701575e-05, + "loss": 0.4104, + "num_input_tokens_seen": 3939528, + "step": 14705 + }, + { + "epoch": 3.855832241153342, + "grad_norm": 0.7473147511482239, + "learning_rate": 3.857518368826157e-05, + "loss": 0.4353, + "num_input_tokens_seen": 3940888, + "step": 14710 + }, + { + "epoch": 3.857142857142857, + "grad_norm": 0.556910514831543, + "learning_rate": 3.856557806255908e-05, + "loss": 0.3997, + "num_input_tokens_seen": 3942296, + "step": 14715 + }, + { + "epoch": 3.8584534731323723, + "grad_norm": 1.1942552328109741, + "learning_rate": 3.8555969597604536e-05, + "loss": 0.3623, + "num_input_tokens_seen": 3943432, + "step": 14720 + }, + { + "epoch": 3.8597640891218874, + "grad_norm": 0.9112830758094788, + "learning_rate": 3.854635829540898e-05, + "loss": 0.2694, + "num_input_tokens_seen": 3944760, + "step": 14725 + }, + { + "epoch": 3.8610747051114025, + "grad_norm": 0.6126102209091187, + "learning_rate": 3.853674415798404e-05, + "loss": 0.2898, + "num_input_tokens_seen": 3947544, + "step": 14730 + }, + { + "epoch": 3.8623853211009176, + "grad_norm": 0.777836263179779, + "learning_rate": 3.8527127187341936e-05, + "loss": 0.3183, + "num_input_tokens_seen": 3949016, + "step": 14735 + }, + { + "epoch": 3.8636959370904327, + "grad_norm": 1.2197712659835815, + "learning_rate": 3.8517507385495486e-05, + "loss": 0.4462, + "num_input_tokens_seen": 3950584, + "step": 14740 + }, + { + "epoch": 3.8650065530799473, + "grad_norm": 0.5715994238853455, + "learning_rate": 3.85078847544581e-05, + "loss": 0.5961, + "num_input_tokens_seen": 3951720, + "step": 14745 + }, + { + "epoch": 3.866317169069463, + "grad_norm": 0.6886200308799744, + "learning_rate": 3.849825929624377e-05, + "loss": 0.315, + "num_input_tokens_seen": 3952984, + "step": 14750 + }, + { + "epoch": 3.8676277850589775, + "grad_norm": 0.7458993792533875, + "learning_rate": 3.8488631012867095e-05, + "loss": 0.4356, + "num_input_tokens_seen": 3954680, + "step": 14755 + }, + { + "epoch": 3.8689384010484926, + "grad_norm": 0.6925996541976929, + "learning_rate": 3.847899990634326e-05, + "loss": 0.3284, + "num_input_tokens_seen": 3956024, + "step": 14760 + }, + { + "epoch": 3.8702490170380077, + "grad_norm": 0.6618605852127075, + "learning_rate": 3.846936597868802e-05, + "loss": 0.4001, + "num_input_tokens_seen": 3957320, + "step": 14765 + }, + { + "epoch": 3.871559633027523, + "grad_norm": 1.2284647226333618, + "learning_rate": 3.845972923191776e-05, + "loss": 0.5437, + "num_input_tokens_seen": 3958376, + "step": 14770 + }, + { + "epoch": 3.872870249017038, + "grad_norm": 0.96384197473526, + "learning_rate": 3.845008966804944e-05, + "loss": 0.3694, + "num_input_tokens_seen": 3959624, + "step": 14775 + }, + { + "epoch": 3.874180865006553, + "grad_norm": 0.999944269657135, + "learning_rate": 3.844044728910058e-05, + "loss": 0.3664, + "num_input_tokens_seen": 3960760, + "step": 14780 + }, + { + "epoch": 3.875491480996068, + "grad_norm": 1.2400370836257935, + "learning_rate": 3.843080209708933e-05, + "loss": 0.3493, + "num_input_tokens_seen": 3962504, + "step": 14785 + }, + { + "epoch": 3.876802096985583, + "grad_norm": 0.5095767974853516, + "learning_rate": 3.8421154094034404e-05, + "loss": 0.3367, + "num_input_tokens_seen": 3964200, + "step": 14790 + }, + { + "epoch": 3.8781127129750983, + "grad_norm": 0.7312006950378418, + "learning_rate": 3.841150328195512e-05, + "loss": 0.3157, + "num_input_tokens_seen": 3965192, + "step": 14795 + }, + { + "epoch": 3.8794233289646134, + "grad_norm": 1.425891399383545, + "learning_rate": 3.840184966287137e-05, + "loss": 0.4685, + "num_input_tokens_seen": 3966744, + "step": 14800 + }, + { + "epoch": 3.8807339449541285, + "grad_norm": 0.6799001693725586, + "learning_rate": 3.839219323880365e-05, + "loss": 0.3698, + "num_input_tokens_seen": 3968440, + "step": 14805 + }, + { + "epoch": 3.8820445609436436, + "grad_norm": 0.942887544631958, + "learning_rate": 3.838253401177302e-05, + "loss": 0.4473, + "num_input_tokens_seen": 3970072, + "step": 14810 + }, + { + "epoch": 3.8833551769331587, + "grad_norm": 0.4990319609642029, + "learning_rate": 3.837287198380114e-05, + "loss": 0.3471, + "num_input_tokens_seen": 3971112, + "step": 14815 + }, + { + "epoch": 3.8846657929226738, + "grad_norm": 8.413534164428711, + "learning_rate": 3.836320715691027e-05, + "loss": 0.4592, + "num_input_tokens_seen": 3972120, + "step": 14820 + }, + { + "epoch": 3.885976408912189, + "grad_norm": 1.3400013446807861, + "learning_rate": 3.835353953312322e-05, + "loss": 0.3803, + "num_input_tokens_seen": 3973448, + "step": 14825 + }, + { + "epoch": 3.8872870249017035, + "grad_norm": 0.508967936038971, + "learning_rate": 3.8343869114463424e-05, + "loss": 0.566, + "num_input_tokens_seen": 3975192, + "step": 14830 + }, + { + "epoch": 3.888597640891219, + "grad_norm": 1.322320580482483, + "learning_rate": 3.8334195902954885e-05, + "loss": 0.458, + "num_input_tokens_seen": 3976344, + "step": 14835 + }, + { + "epoch": 3.8899082568807337, + "grad_norm": 0.6365997791290283, + "learning_rate": 3.8324519900622175e-05, + "loss": 0.4071, + "num_input_tokens_seen": 3977832, + "step": 14840 + }, + { + "epoch": 3.8912188728702493, + "grad_norm": 0.9568093419075012, + "learning_rate": 3.8314841109490474e-05, + "loss": 0.3831, + "num_input_tokens_seen": 3979128, + "step": 14845 + }, + { + "epoch": 3.892529488859764, + "grad_norm": 1.2392022609710693, + "learning_rate": 3.830515953158553e-05, + "loss": 0.429, + "num_input_tokens_seen": 3980248, + "step": 14850 + }, + { + "epoch": 3.893840104849279, + "grad_norm": 1.2005584239959717, + "learning_rate": 3.82954751689337e-05, + "loss": 0.3152, + "num_input_tokens_seen": 3981320, + "step": 14855 + }, + { + "epoch": 3.895150720838794, + "grad_norm": 0.9969881772994995, + "learning_rate": 3.828578802356188e-05, + "loss": 0.5651, + "num_input_tokens_seen": 3982584, + "step": 14860 + }, + { + "epoch": 3.896461336828309, + "grad_norm": 1.0862873792648315, + "learning_rate": 3.8276098097497584e-05, + "loss": 0.3039, + "num_input_tokens_seen": 3983928, + "step": 14865 + }, + { + "epoch": 3.8977719528178243, + "grad_norm": 0.5690191984176636, + "learning_rate": 3.8266405392768904e-05, + "loss": 0.3963, + "num_input_tokens_seen": 3985816, + "step": 14870 + }, + { + "epoch": 3.8990825688073394, + "grad_norm": 0.9804603457450867, + "learning_rate": 3.8256709911404484e-05, + "loss": 0.4003, + "num_input_tokens_seen": 3987208, + "step": 14875 + }, + { + "epoch": 3.9003931847968545, + "grad_norm": 0.8451335430145264, + "learning_rate": 3.82470116554336e-05, + "loss": 0.4305, + "num_input_tokens_seen": 3988440, + "step": 14880 + }, + { + "epoch": 3.9017038007863696, + "grad_norm": 1.817413091659546, + "learning_rate": 3.823731062688605e-05, + "loss": 0.3642, + "num_input_tokens_seen": 3989464, + "step": 14885 + }, + { + "epoch": 3.9030144167758847, + "grad_norm": 0.9638396501541138, + "learning_rate": 3.8227606827792265e-05, + "loss": 0.3803, + "num_input_tokens_seen": 3990872, + "step": 14890 + }, + { + "epoch": 3.9043250327653998, + "grad_norm": 1.1120113134384155, + "learning_rate": 3.8217900260183224e-05, + "loss": 0.3857, + "num_input_tokens_seen": 3992008, + "step": 14895 + }, + { + "epoch": 3.905635648754915, + "grad_norm": 0.621444046497345, + "learning_rate": 3.82081909260905e-05, + "loss": 0.282, + "num_input_tokens_seen": 3993288, + "step": 14900 + }, + { + "epoch": 3.90694626474443, + "grad_norm": 0.5280365347862244, + "learning_rate": 3.819847882754623e-05, + "loss": 0.3951, + "num_input_tokens_seen": 3994728, + "step": 14905 + }, + { + "epoch": 3.908256880733945, + "grad_norm": 0.5428124070167542, + "learning_rate": 3.8188763966583165e-05, + "loss": 0.3428, + "num_input_tokens_seen": 3995928, + "step": 14910 + }, + { + "epoch": 3.90956749672346, + "grad_norm": 0.9505031108856201, + "learning_rate": 3.817904634523458e-05, + "loss": 0.4426, + "num_input_tokens_seen": 3997208, + "step": 14915 + }, + { + "epoch": 3.9108781127129753, + "grad_norm": 1.3737047910690308, + "learning_rate": 3.816932596553436e-05, + "loss": 0.3621, + "num_input_tokens_seen": 3998808, + "step": 14920 + }, + { + "epoch": 3.91218872870249, + "grad_norm": 0.6429957747459412, + "learning_rate": 3.8159602829516975e-05, + "loss": 0.4339, + "num_input_tokens_seen": 4000232, + "step": 14925 + }, + { + "epoch": 3.9134993446920054, + "grad_norm": 1.4048848152160645, + "learning_rate": 3.8149876939217454e-05, + "loss": 0.3655, + "num_input_tokens_seen": 4001256, + "step": 14930 + }, + { + "epoch": 3.91480996068152, + "grad_norm": 1.26173734664917, + "learning_rate": 3.814014829667142e-05, + "loss": 0.3242, + "num_input_tokens_seen": 4002568, + "step": 14935 + }, + { + "epoch": 3.9161205766710356, + "grad_norm": 0.738139808177948, + "learning_rate": 3.8130416903915045e-05, + "loss": 0.3969, + "num_input_tokens_seen": 4003976, + "step": 14940 + }, + { + "epoch": 3.9174311926605503, + "grad_norm": 0.8491547703742981, + "learning_rate": 3.812068276298509e-05, + "loss": 0.3615, + "num_input_tokens_seen": 4005848, + "step": 14945 + }, + { + "epoch": 3.9187418086500654, + "grad_norm": 1.571791648864746, + "learning_rate": 3.811094587591892e-05, + "loss": 0.3665, + "num_input_tokens_seen": 4006808, + "step": 14950 + }, + { + "epoch": 3.9200524246395805, + "grad_norm": 0.9408077001571655, + "learning_rate": 3.810120624475443e-05, + "loss": 0.3468, + "num_input_tokens_seen": 4008184, + "step": 14955 + }, + { + "epoch": 3.9213630406290956, + "grad_norm": 0.4982743263244629, + "learning_rate": 3.809146387153011e-05, + "loss": 0.3887, + "num_input_tokens_seen": 4009656, + "step": 14960 + }, + { + "epoch": 3.9226736566186107, + "grad_norm": 0.7377883791923523, + "learning_rate": 3.808171875828501e-05, + "loss": 0.5241, + "num_input_tokens_seen": 4010760, + "step": 14965 + }, + { + "epoch": 3.9239842726081258, + "grad_norm": 2.26383113861084, + "learning_rate": 3.8071970907058786e-05, + "loss": 0.6519, + "num_input_tokens_seen": 4011752, + "step": 14970 + }, + { + "epoch": 3.925294888597641, + "grad_norm": 0.7709552645683289, + "learning_rate": 3.806222031989164e-05, + "loss": 0.4477, + "num_input_tokens_seen": 4012872, + "step": 14975 + }, + { + "epoch": 3.926605504587156, + "grad_norm": 0.672200083732605, + "learning_rate": 3.8052466998824344e-05, + "loss": 0.2941, + "num_input_tokens_seen": 4013896, + "step": 14980 + }, + { + "epoch": 3.927916120576671, + "grad_norm": 0.6008850932121277, + "learning_rate": 3.804271094589826e-05, + "loss": 0.4397, + "num_input_tokens_seen": 4015560, + "step": 14985 + }, + { + "epoch": 3.929226736566186, + "grad_norm": 1.8690969944000244, + "learning_rate": 3.803295216315532e-05, + "loss": 0.3717, + "num_input_tokens_seen": 4016696, + "step": 14990 + }, + { + "epoch": 3.9305373525557012, + "grad_norm": 1.126692771911621, + "learning_rate": 3.802319065263801e-05, + "loss": 0.3805, + "num_input_tokens_seen": 4017800, + "step": 14995 + }, + { + "epoch": 3.9318479685452163, + "grad_norm": 2.965583324432373, + "learning_rate": 3.8013426416389385e-05, + "loss": 0.3815, + "num_input_tokens_seen": 4019224, + "step": 15000 + }, + { + "epoch": 3.9331585845347314, + "grad_norm": 1.589913010597229, + "learning_rate": 3.80036594564531e-05, + "loss": 0.3814, + "num_input_tokens_seen": 4020488, + "step": 15005 + }, + { + "epoch": 3.9344692005242465, + "grad_norm": 1.0554085969924927, + "learning_rate": 3.799388977487337e-05, + "loss": 0.3393, + "num_input_tokens_seen": 4021592, + "step": 15010 + }, + { + "epoch": 3.9357798165137616, + "grad_norm": 0.9072408080101013, + "learning_rate": 3.798411737369495e-05, + "loss": 0.3479, + "num_input_tokens_seen": 4022728, + "step": 15015 + }, + { + "epoch": 3.9370904325032763, + "grad_norm": 0.8650107979774475, + "learning_rate": 3.7974342254963214e-05, + "loss": 0.7378, + "num_input_tokens_seen": 4024328, + "step": 15020 + }, + { + "epoch": 3.938401048492792, + "grad_norm": 0.5100976228713989, + "learning_rate": 3.7964564420724044e-05, + "loss": 0.4576, + "num_input_tokens_seen": 4025880, + "step": 15025 + }, + { + "epoch": 3.9397116644823065, + "grad_norm": 0.6682380437850952, + "learning_rate": 3.7954783873023955e-05, + "loss": 0.725, + "num_input_tokens_seen": 4027192, + "step": 15030 + }, + { + "epoch": 3.941022280471822, + "grad_norm": 2.240804433822632, + "learning_rate": 3.7945000613909975e-05, + "loss": 0.3187, + "num_input_tokens_seen": 4028536, + "step": 15035 + }, + { + "epoch": 3.9423328964613367, + "grad_norm": 3.352780818939209, + "learning_rate": 3.793521464542974e-05, + "loss": 0.4039, + "num_input_tokens_seen": 4029944, + "step": 15040 + }, + { + "epoch": 3.9436435124508518, + "grad_norm": 1.2332531213760376, + "learning_rate": 3.7925425969631435e-05, + "loss": 0.473, + "num_input_tokens_seen": 4031080, + "step": 15045 + }, + { + "epoch": 3.944954128440367, + "grad_norm": 0.6892217993736267, + "learning_rate": 3.7915634588563804e-05, + "loss": 0.3016, + "num_input_tokens_seen": 4033128, + "step": 15050 + }, + { + "epoch": 3.946264744429882, + "grad_norm": 1.0183178186416626, + "learning_rate": 3.790584050427616e-05, + "loss": 0.4441, + "num_input_tokens_seen": 4034472, + "step": 15055 + }, + { + "epoch": 3.947575360419397, + "grad_norm": 0.49372345209121704, + "learning_rate": 3.7896043718818406e-05, + "loss": 0.4492, + "num_input_tokens_seen": 4035768, + "step": 15060 + }, + { + "epoch": 3.948885976408912, + "grad_norm": 1.1246888637542725, + "learning_rate": 3.788624423424099e-05, + "loss": 0.4784, + "num_input_tokens_seen": 4036984, + "step": 15065 + }, + { + "epoch": 3.9501965923984272, + "grad_norm": 1.5061033964157104, + "learning_rate": 3.7876442052594915e-05, + "loss": 0.7114, + "num_input_tokens_seen": 4038184, + "step": 15070 + }, + { + "epoch": 3.9515072083879423, + "grad_norm": 0.8146348595619202, + "learning_rate": 3.7866637175931774e-05, + "loss": 0.3764, + "num_input_tokens_seen": 4039448, + "step": 15075 + }, + { + "epoch": 3.9528178243774574, + "grad_norm": 0.6819890141487122, + "learning_rate": 3.7856829606303696e-05, + "loss": 0.4428, + "num_input_tokens_seen": 4040792, + "step": 15080 + }, + { + "epoch": 3.9541284403669725, + "grad_norm": 0.9135717153549194, + "learning_rate": 3.784701934576339e-05, + "loss": 0.4042, + "num_input_tokens_seen": 4042008, + "step": 15085 + }, + { + "epoch": 3.9554390563564876, + "grad_norm": 0.33685848116874695, + "learning_rate": 3.783720639636415e-05, + "loss": 0.3299, + "num_input_tokens_seen": 4043368, + "step": 15090 + }, + { + "epoch": 3.9567496723460027, + "grad_norm": 0.48980283737182617, + "learning_rate": 3.782739076015978e-05, + "loss": 0.4979, + "num_input_tokens_seen": 4044472, + "step": 15095 + }, + { + "epoch": 3.958060288335518, + "grad_norm": 2.2941341400146484, + "learning_rate": 3.7817572439204695e-05, + "loss": 0.2946, + "num_input_tokens_seen": 4045400, + "step": 15100 + }, + { + "epoch": 3.959370904325033, + "grad_norm": 1.2184138298034668, + "learning_rate": 3.780775143555384e-05, + "loss": 0.3135, + "num_input_tokens_seen": 4046760, + "step": 15105 + }, + { + "epoch": 3.960681520314548, + "grad_norm": 0.3761690855026245, + "learning_rate": 3.779792775126275e-05, + "loss": 0.4704, + "num_input_tokens_seen": 4048360, + "step": 15110 + }, + { + "epoch": 3.9619921363040627, + "grad_norm": 1.7997961044311523, + "learning_rate": 3.778810138838748e-05, + "loss": 0.3108, + "num_input_tokens_seen": 4049528, + "step": 15115 + }, + { + "epoch": 3.963302752293578, + "grad_norm": 1.2706730365753174, + "learning_rate": 3.777827234898469e-05, + "loss": 0.5345, + "num_input_tokens_seen": 4050664, + "step": 15120 + }, + { + "epoch": 3.964613368283093, + "grad_norm": 0.5991653203964233, + "learning_rate": 3.776844063511158e-05, + "loss": 0.4709, + "num_input_tokens_seen": 4052216, + "step": 15125 + }, + { + "epoch": 3.9659239842726084, + "grad_norm": 0.5916756987571716, + "learning_rate": 3.7758606248825914e-05, + "loss": 0.4766, + "num_input_tokens_seen": 4053720, + "step": 15130 + }, + { + "epoch": 3.967234600262123, + "grad_norm": 1.0458310842514038, + "learning_rate": 3.774876919218599e-05, + "loss": 0.599, + "num_input_tokens_seen": 4055080, + "step": 15135 + }, + { + "epoch": 3.968545216251638, + "grad_norm": 0.6298121213912964, + "learning_rate": 3.773892946725071e-05, + "loss": 0.3467, + "num_input_tokens_seen": 4056328, + "step": 15140 + }, + { + "epoch": 3.9698558322411532, + "grad_norm": 0.512183427810669, + "learning_rate": 3.7729087076079505e-05, + "loss": 0.4805, + "num_input_tokens_seen": 4057944, + "step": 15145 + }, + { + "epoch": 3.9711664482306683, + "grad_norm": 0.834551990032196, + "learning_rate": 3.771924202073236e-05, + "loss": 0.4119, + "num_input_tokens_seen": 4059128, + "step": 15150 + }, + { + "epoch": 3.9724770642201834, + "grad_norm": 0.8493472337722778, + "learning_rate": 3.770939430326985e-05, + "loss": 0.3223, + "num_input_tokens_seen": 4060472, + "step": 15155 + }, + { + "epoch": 3.9737876802096985, + "grad_norm": 0.6324858665466309, + "learning_rate": 3.7699543925753064e-05, + "loss": 0.3399, + "num_input_tokens_seen": 4061608, + "step": 15160 + }, + { + "epoch": 3.9750982961992136, + "grad_norm": 0.6789525151252747, + "learning_rate": 3.768969089024368e-05, + "loss": 0.4196, + "num_input_tokens_seen": 4063624, + "step": 15165 + }, + { + "epoch": 3.9764089121887287, + "grad_norm": 0.9688178300857544, + "learning_rate": 3.767983519880392e-05, + "loss": 0.3225, + "num_input_tokens_seen": 4066152, + "step": 15170 + }, + { + "epoch": 3.977719528178244, + "grad_norm": 2.5714974403381348, + "learning_rate": 3.766997685349656e-05, + "loss": 0.4197, + "num_input_tokens_seen": 4067304, + "step": 15175 + }, + { + "epoch": 3.979030144167759, + "grad_norm": 0.7018616795539856, + "learning_rate": 3.766011585638494e-05, + "loss": 0.3074, + "num_input_tokens_seen": 4068456, + "step": 15180 + }, + { + "epoch": 3.980340760157274, + "grad_norm": 0.8959025144577026, + "learning_rate": 3.765025220953294e-05, + "loss": 0.2531, + "num_input_tokens_seen": 4069864, + "step": 15185 + }, + { + "epoch": 3.981651376146789, + "grad_norm": 0.8549515604972839, + "learning_rate": 3.764038591500502e-05, + "loss": 0.3422, + "num_input_tokens_seen": 4071096, + "step": 15190 + }, + { + "epoch": 3.982961992136304, + "grad_norm": 1.257409930229187, + "learning_rate": 3.7630516974866166e-05, + "loss": 0.6255, + "num_input_tokens_seen": 4072504, + "step": 15195 + }, + { + "epoch": 3.9842726081258193, + "grad_norm": 0.9116142988204956, + "learning_rate": 3.7620645391181926e-05, + "loss": 0.3248, + "num_input_tokens_seen": 4073880, + "step": 15200 + }, + { + "epoch": 3.9855832241153344, + "grad_norm": 0.2956502437591553, + "learning_rate": 3.761077116601842e-05, + "loss": 0.38, + "num_input_tokens_seen": 4075096, + "step": 15205 + }, + { + "epoch": 3.986893840104849, + "grad_norm": 0.8169086575508118, + "learning_rate": 3.760089430144229e-05, + "loss": 0.6168, + "num_input_tokens_seen": 4076648, + "step": 15210 + }, + { + "epoch": 3.9882044560943646, + "grad_norm": 1.2878860235214233, + "learning_rate": 3.759101479952076e-05, + "loss": 0.4509, + "num_input_tokens_seen": 4078088, + "step": 15215 + }, + { + "epoch": 3.9895150720838792, + "grad_norm": 0.4051441252231598, + "learning_rate": 3.7581132662321597e-05, + "loss": 0.4403, + "num_input_tokens_seen": 4079416, + "step": 15220 + }, + { + "epoch": 3.9908256880733948, + "grad_norm": 1.551731824874878, + "learning_rate": 3.7571247891913095e-05, + "loss": 0.3731, + "num_input_tokens_seen": 4081288, + "step": 15225 + }, + { + "epoch": 3.9921363040629094, + "grad_norm": 1.1671833992004395, + "learning_rate": 3.756136049036413e-05, + "loss": 0.5849, + "num_input_tokens_seen": 4082584, + "step": 15230 + }, + { + "epoch": 3.9934469200524245, + "grad_norm": 0.7103251814842224, + "learning_rate": 3.7551470459744126e-05, + "loss": 0.3505, + "num_input_tokens_seen": 4083848, + "step": 15235 + }, + { + "epoch": 3.9947575360419396, + "grad_norm": 1.4447342157363892, + "learning_rate": 3.754157780212304e-05, + "loss": 0.4316, + "num_input_tokens_seen": 4085160, + "step": 15240 + }, + { + "epoch": 3.9960681520314547, + "grad_norm": 0.3061644434928894, + "learning_rate": 3.753168251957139e-05, + "loss": 0.2167, + "num_input_tokens_seen": 4086600, + "step": 15245 + }, + { + "epoch": 3.99737876802097, + "grad_norm": 0.6745178699493408, + "learning_rate": 3.752178461416024e-05, + "loss": 0.316, + "num_input_tokens_seen": 4087736, + "step": 15250 + }, + { + "epoch": 3.998689384010485, + "grad_norm": 2.4452922344207764, + "learning_rate": 3.75118840879612e-05, + "loss": 0.4972, + "num_input_tokens_seen": 4089000, + "step": 15255 + }, + { + "epoch": 4.0, + "grad_norm": 1.078931450843811, + "learning_rate": 3.750198094304644e-05, + "loss": 0.5023, + "num_input_tokens_seen": 4089872, + "step": 15260 + }, + { + "epoch": 4.001048492791612, + "eval_loss": 0.5102683901786804, + "eval_runtime": 16.8609, + "eval_samples_per_second": 50.294, + "eval_steps_per_second": 25.147, + "num_input_tokens_seen": 4090736, + "step": 15264 + }, + { + "epoch": 4.001310615989515, + "grad_norm": 0.546513020992279, + "learning_rate": 3.749207518148869e-05, + "loss": 0.4168, + "num_input_tokens_seen": 4091024, + "step": 15265 + }, + { + "epoch": 4.00262123197903, + "grad_norm": 1.7809734344482422, + "learning_rate": 3.7482166805361175e-05, + "loss": 0.2953, + "num_input_tokens_seen": 4092256, + "step": 15270 + }, + { + "epoch": 4.003931847968545, + "grad_norm": 2.215810775756836, + "learning_rate": 3.747225581673771e-05, + "loss": 0.2694, + "num_input_tokens_seen": 4093168, + "step": 15275 + }, + { + "epoch": 4.00524246395806, + "grad_norm": 0.3976542353630066, + "learning_rate": 3.746234221769266e-05, + "loss": 0.3101, + "num_input_tokens_seen": 4094480, + "step": 15280 + }, + { + "epoch": 4.006553079947575, + "grad_norm": 0.8934128880500793, + "learning_rate": 3.7452426010300915e-05, + "loss": 1.021, + "num_input_tokens_seen": 4095472, + "step": 15285 + }, + { + "epoch": 4.007863695937091, + "grad_norm": 1.4365705251693726, + "learning_rate": 3.7442507196637925e-05, + "loss": 0.4153, + "num_input_tokens_seen": 4096656, + "step": 15290 + }, + { + "epoch": 4.009174311926605, + "grad_norm": 1.148489236831665, + "learning_rate": 3.743258577877968e-05, + "loss": 0.5239, + "num_input_tokens_seen": 4097760, + "step": 15295 + }, + { + "epoch": 4.010484927916121, + "grad_norm": 1.1645634174346924, + "learning_rate": 3.74226617588027e-05, + "loss": 0.274, + "num_input_tokens_seen": 4099168, + "step": 15300 + }, + { + "epoch": 4.011795543905635, + "grad_norm": 0.5938572287559509, + "learning_rate": 3.7412735138784096e-05, + "loss": 0.3281, + "num_input_tokens_seen": 4100576, + "step": 15305 + }, + { + "epoch": 4.013106159895151, + "grad_norm": 0.8798007965087891, + "learning_rate": 3.740280592080147e-05, + "loss": 0.3521, + "num_input_tokens_seen": 4101904, + "step": 15310 + }, + { + "epoch": 4.014416775884666, + "grad_norm": 0.8663793206214905, + "learning_rate": 3.7392874106932985e-05, + "loss": 0.3271, + "num_input_tokens_seen": 4103584, + "step": 15315 + }, + { + "epoch": 4.015727391874181, + "grad_norm": 1.3925068378448486, + "learning_rate": 3.738293969925737e-05, + "loss": 0.3251, + "num_input_tokens_seen": 4104640, + "step": 15320 + }, + { + "epoch": 4.017038007863696, + "grad_norm": 0.7780625820159912, + "learning_rate": 3.737300269985388e-05, + "loss": 0.3175, + "num_input_tokens_seen": 4106272, + "step": 15325 + }, + { + "epoch": 4.018348623853211, + "grad_norm": 0.8553165793418884, + "learning_rate": 3.736306311080229e-05, + "loss": 0.3868, + "num_input_tokens_seen": 4107808, + "step": 15330 + }, + { + "epoch": 4.019659239842726, + "grad_norm": 2.1791326999664307, + "learning_rate": 3.7353120934182956e-05, + "loss": 0.5514, + "num_input_tokens_seen": 4109152, + "step": 15335 + }, + { + "epoch": 4.0209698558322415, + "grad_norm": 0.5489323735237122, + "learning_rate": 3.7343176172076755e-05, + "loss": 0.22, + "num_input_tokens_seen": 4110496, + "step": 15340 + }, + { + "epoch": 4.022280471821756, + "grad_norm": 0.5350883603096008, + "learning_rate": 3.7333228826565115e-05, + "loss": 0.3097, + "num_input_tokens_seen": 4111824, + "step": 15345 + }, + { + "epoch": 4.023591087811272, + "grad_norm": 1.708803653717041, + "learning_rate": 3.7323278899729986e-05, + "loss": 0.3675, + "num_input_tokens_seen": 4112880, + "step": 15350 + }, + { + "epoch": 4.024901703800786, + "grad_norm": 3.5905239582061768, + "learning_rate": 3.731332639365387e-05, + "loss": 0.3797, + "num_input_tokens_seen": 4113840, + "step": 15355 + }, + { + "epoch": 4.026212319790301, + "grad_norm": 0.9438067078590393, + "learning_rate": 3.7303371310419835e-05, + "loss": 0.3725, + "num_input_tokens_seen": 4114800, + "step": 15360 + }, + { + "epoch": 4.027522935779817, + "grad_norm": 1.180009126663208, + "learning_rate": 3.729341365211143e-05, + "loss": 0.391, + "num_input_tokens_seen": 4116416, + "step": 15365 + }, + { + "epoch": 4.028833551769331, + "grad_norm": 1.380281686782837, + "learning_rate": 3.7283453420812786e-05, + "loss": 0.4312, + "num_input_tokens_seen": 4117584, + "step": 15370 + }, + { + "epoch": 4.030144167758847, + "grad_norm": 1.1943459510803223, + "learning_rate": 3.727349061860857e-05, + "loss": 0.4443, + "num_input_tokens_seen": 4118848, + "step": 15375 + }, + { + "epoch": 4.031454783748361, + "grad_norm": 0.8748984336853027, + "learning_rate": 3.726352524758397e-05, + "loss": 0.3164, + "num_input_tokens_seen": 4120032, + "step": 15380 + }, + { + "epoch": 4.032765399737877, + "grad_norm": 0.9514562487602234, + "learning_rate": 3.725355730982474e-05, + "loss": 0.3099, + "num_input_tokens_seen": 4121680, + "step": 15385 + }, + { + "epoch": 4.034076015727392, + "grad_norm": 1.200313925743103, + "learning_rate": 3.724358680741713e-05, + "loss": 0.2075, + "num_input_tokens_seen": 4122800, + "step": 15390 + }, + { + "epoch": 4.035386631716907, + "grad_norm": 0.6990623474121094, + "learning_rate": 3.723361374244795e-05, + "loss": 0.397, + "num_input_tokens_seen": 4124336, + "step": 15395 + }, + { + "epoch": 4.036697247706422, + "grad_norm": 0.7130845189094543, + "learning_rate": 3.722363811700455e-05, + "loss": 0.4406, + "num_input_tokens_seen": 4125904, + "step": 15400 + }, + { + "epoch": 4.038007863695937, + "grad_norm": 2.354475498199463, + "learning_rate": 3.721365993317482e-05, + "loss": 0.3327, + "num_input_tokens_seen": 4127376, + "step": 15405 + }, + { + "epoch": 4.039318479685452, + "grad_norm": 1.180155873298645, + "learning_rate": 3.720367919304717e-05, + "loss": 0.4066, + "num_input_tokens_seen": 4128944, + "step": 15410 + }, + { + "epoch": 4.0406290956749675, + "grad_norm": 0.7764495015144348, + "learning_rate": 3.7193695898710545e-05, + "loss": 0.2773, + "num_input_tokens_seen": 4130176, + "step": 15415 + }, + { + "epoch": 4.041939711664482, + "grad_norm": 1.026907205581665, + "learning_rate": 3.718371005225445e-05, + "loss": 0.3127, + "num_input_tokens_seen": 4131680, + "step": 15420 + }, + { + "epoch": 4.043250327653998, + "grad_norm": 1.0260189771652222, + "learning_rate": 3.717372165576888e-05, + "loss": 0.3635, + "num_input_tokens_seen": 4132736, + "step": 15425 + }, + { + "epoch": 4.044560943643512, + "grad_norm": 1.6111130714416504, + "learning_rate": 3.71637307113444e-05, + "loss": 0.2235, + "num_input_tokens_seen": 4134240, + "step": 15430 + }, + { + "epoch": 4.045871559633028, + "grad_norm": 1.3560514450073242, + "learning_rate": 3.715373722107211e-05, + "loss": 0.3859, + "num_input_tokens_seen": 4135616, + "step": 15435 + }, + { + "epoch": 4.047182175622543, + "grad_norm": 1.0535069704055786, + "learning_rate": 3.714374118704362e-05, + "loss": 0.3887, + "num_input_tokens_seen": 4137376, + "step": 15440 + }, + { + "epoch": 4.048492791612058, + "grad_norm": 1.3934121131896973, + "learning_rate": 3.713374261135107e-05, + "loss": 0.5808, + "num_input_tokens_seen": 4138640, + "step": 15445 + }, + { + "epoch": 4.049803407601573, + "grad_norm": 1.4915348291397095, + "learning_rate": 3.712374149608717e-05, + "loss": 0.3088, + "num_input_tokens_seen": 4140032, + "step": 15450 + }, + { + "epoch": 4.051114023591087, + "grad_norm": 1.4573289155960083, + "learning_rate": 3.711373784334511e-05, + "loss": 0.5765, + "num_input_tokens_seen": 4141488, + "step": 15455 + }, + { + "epoch": 4.052424639580603, + "grad_norm": 1.299834966659546, + "learning_rate": 3.7103731655218664e-05, + "loss": 0.5041, + "num_input_tokens_seen": 4142864, + "step": 15460 + }, + { + "epoch": 4.053735255570118, + "grad_norm": 1.0833899974822998, + "learning_rate": 3.7093722933802095e-05, + "loss": 0.2927, + "num_input_tokens_seen": 4145888, + "step": 15465 + }, + { + "epoch": 4.055045871559633, + "grad_norm": 1.2428661584854126, + "learning_rate": 3.70837116811902e-05, + "loss": 0.3363, + "num_input_tokens_seen": 4147264, + "step": 15470 + }, + { + "epoch": 4.056356487549148, + "grad_norm": 0.4387124180793762, + "learning_rate": 3.707369789947834e-05, + "loss": 0.3501, + "num_input_tokens_seen": 4149184, + "step": 15475 + }, + { + "epoch": 4.057667103538663, + "grad_norm": 2.007150411605835, + "learning_rate": 3.706368159076238e-05, + "loss": 0.4669, + "num_input_tokens_seen": 4150624, + "step": 15480 + }, + { + "epoch": 4.058977719528178, + "grad_norm": 3.2900123596191406, + "learning_rate": 3.7053662757138686e-05, + "loss": 0.378, + "num_input_tokens_seen": 4152256, + "step": 15485 + }, + { + "epoch": 4.0602883355176935, + "grad_norm": 0.9015100002288818, + "learning_rate": 3.704364140070421e-05, + "loss": 0.3569, + "num_input_tokens_seen": 4153552, + "step": 15490 + }, + { + "epoch": 4.061598951507208, + "grad_norm": 1.2322052717208862, + "learning_rate": 3.703361752355641e-05, + "loss": 0.3556, + "num_input_tokens_seen": 4154800, + "step": 15495 + }, + { + "epoch": 4.062909567496724, + "grad_norm": 0.9732781648635864, + "learning_rate": 3.702359112779325e-05, + "loss": 0.2273, + "num_input_tokens_seen": 4155872, + "step": 15500 + }, + { + "epoch": 4.064220183486238, + "grad_norm": 1.3891141414642334, + "learning_rate": 3.701356221551324e-05, + "loss": 0.3922, + "num_input_tokens_seen": 4157200, + "step": 15505 + }, + { + "epoch": 4.065530799475754, + "grad_norm": 0.87093186378479, + "learning_rate": 3.700353078881541e-05, + "loss": 0.2537, + "num_input_tokens_seen": 4158416, + "step": 15510 + }, + { + "epoch": 4.066841415465269, + "grad_norm": 0.614954948425293, + "learning_rate": 3.699349684979932e-05, + "loss": 0.3834, + "num_input_tokens_seen": 4159888, + "step": 15515 + }, + { + "epoch": 4.068152031454784, + "grad_norm": 0.8495243191719055, + "learning_rate": 3.6983460400565086e-05, + "loss": 0.1872, + "num_input_tokens_seen": 4161120, + "step": 15520 + }, + { + "epoch": 4.069462647444299, + "grad_norm": 2.0463080406188965, + "learning_rate": 3.697342144321329e-05, + "loss": 0.6093, + "num_input_tokens_seen": 4162288, + "step": 15525 + }, + { + "epoch": 4.070773263433814, + "grad_norm": 0.9853081703186035, + "learning_rate": 3.6963379979845066e-05, + "loss": 0.4939, + "num_input_tokens_seen": 4165056, + "step": 15530 + }, + { + "epoch": 4.072083879423329, + "grad_norm": 1.3450438976287842, + "learning_rate": 3.6953336012562084e-05, + "loss": 0.363, + "num_input_tokens_seen": 4166656, + "step": 15535 + }, + { + "epoch": 4.073394495412844, + "grad_norm": 1.042601466178894, + "learning_rate": 3.6943289543466536e-05, + "loss": 0.2511, + "num_input_tokens_seen": 4167728, + "step": 15540 + }, + { + "epoch": 4.074705111402359, + "grad_norm": 0.9710046052932739, + "learning_rate": 3.693324057466111e-05, + "loss": 0.4394, + "num_input_tokens_seen": 4169120, + "step": 15545 + }, + { + "epoch": 4.076015727391874, + "grad_norm": 0.7509154677391052, + "learning_rate": 3.692318910824906e-05, + "loss": 0.2652, + "num_input_tokens_seen": 4170864, + "step": 15550 + }, + { + "epoch": 4.077326343381389, + "grad_norm": 0.9625009894371033, + "learning_rate": 3.691313514633413e-05, + "loss": 0.4412, + "num_input_tokens_seen": 4171824, + "step": 15555 + }, + { + "epoch": 4.078636959370904, + "grad_norm": 0.9617549777030945, + "learning_rate": 3.69030786910206e-05, + "loss": 0.3547, + "num_input_tokens_seen": 4173056, + "step": 15560 + }, + { + "epoch": 4.0799475753604195, + "grad_norm": 0.7804156541824341, + "learning_rate": 3.6893019744413265e-05, + "loss": 0.2977, + "num_input_tokens_seen": 4174080, + "step": 15565 + }, + { + "epoch": 4.081258191349934, + "grad_norm": 1.0347810983657837, + "learning_rate": 3.688295830861744e-05, + "loss": 0.3341, + "num_input_tokens_seen": 4175136, + "step": 15570 + }, + { + "epoch": 4.08256880733945, + "grad_norm": 0.8177746534347534, + "learning_rate": 3.687289438573899e-05, + "loss": 0.2839, + "num_input_tokens_seen": 4176208, + "step": 15575 + }, + { + "epoch": 4.083879423328964, + "grad_norm": 0.5250045657157898, + "learning_rate": 3.6862827977884246e-05, + "loss": 0.3901, + "num_input_tokens_seen": 4177680, + "step": 15580 + }, + { + "epoch": 4.08519003931848, + "grad_norm": 1.6613296270370483, + "learning_rate": 3.6852759087160105e-05, + "loss": 0.339, + "num_input_tokens_seen": 4179632, + "step": 15585 + }, + { + "epoch": 4.086500655307995, + "grad_norm": 1.1001238822937012, + "learning_rate": 3.6842687715673977e-05, + "loss": 0.3367, + "num_input_tokens_seen": 4180992, + "step": 15590 + }, + { + "epoch": 4.08781127129751, + "grad_norm": 1.2594919204711914, + "learning_rate": 3.683261386553376e-05, + "loss": 0.3727, + "num_input_tokens_seen": 4182608, + "step": 15595 + }, + { + "epoch": 4.089121887287025, + "grad_norm": 0.889535665512085, + "learning_rate": 3.6822537538847914e-05, + "loss": 0.3456, + "num_input_tokens_seen": 4183952, + "step": 15600 + }, + { + "epoch": 4.09043250327654, + "grad_norm": 1.9227426052093506, + "learning_rate": 3.6812458737725384e-05, + "loss": 0.3727, + "num_input_tokens_seen": 4185920, + "step": 15605 + }, + { + "epoch": 4.091743119266055, + "grad_norm": 2.039670705795288, + "learning_rate": 3.680237746427565e-05, + "loss": 0.3772, + "num_input_tokens_seen": 4186896, + "step": 15610 + }, + { + "epoch": 4.0930537352555705, + "grad_norm": 0.8021466732025146, + "learning_rate": 3.679229372060871e-05, + "loss": 0.3554, + "num_input_tokens_seen": 4188032, + "step": 15615 + }, + { + "epoch": 4.094364351245085, + "grad_norm": 0.624109148979187, + "learning_rate": 3.678220750883507e-05, + "loss": 0.2968, + "num_input_tokens_seen": 4189728, + "step": 15620 + }, + { + "epoch": 4.095674967234601, + "grad_norm": 1.532147765159607, + "learning_rate": 3.6772118831065755e-05, + "loss": 0.4494, + "num_input_tokens_seen": 4191024, + "step": 15625 + }, + { + "epoch": 4.096985583224115, + "grad_norm": 1.4903333187103271, + "learning_rate": 3.6762027689412305e-05, + "loss": 0.5334, + "num_input_tokens_seen": 4192672, + "step": 15630 + }, + { + "epoch": 4.09829619921363, + "grad_norm": 0.9615368247032166, + "learning_rate": 3.675193408598679e-05, + "loss": 0.435, + "num_input_tokens_seen": 4194144, + "step": 15635 + }, + { + "epoch": 4.0996068152031455, + "grad_norm": 0.8903807401657104, + "learning_rate": 3.674183802290178e-05, + "loss": 0.3627, + "num_input_tokens_seen": 4196128, + "step": 15640 + }, + { + "epoch": 4.10091743119266, + "grad_norm": 1.0330955982208252, + "learning_rate": 3.6731739502270344e-05, + "loss": 0.3727, + "num_input_tokens_seen": 4197856, + "step": 15645 + }, + { + "epoch": 4.102228047182176, + "grad_norm": 0.718490719795227, + "learning_rate": 3.6721638526206115e-05, + "loss": 0.3507, + "num_input_tokens_seen": 4199008, + "step": 15650 + }, + { + "epoch": 4.10353866317169, + "grad_norm": 0.49733874201774597, + "learning_rate": 3.671153509682319e-05, + "loss": 0.3323, + "num_input_tokens_seen": 4200752, + "step": 15655 + }, + { + "epoch": 4.104849279161206, + "grad_norm": 0.9984380006790161, + "learning_rate": 3.67014292162362e-05, + "loss": 0.4535, + "num_input_tokens_seen": 4201888, + "step": 15660 + }, + { + "epoch": 4.1061598951507206, + "grad_norm": 0.9477693438529968, + "learning_rate": 3.66913208865603e-05, + "loss": 0.291, + "num_input_tokens_seen": 4202928, + "step": 15665 + }, + { + "epoch": 4.107470511140236, + "grad_norm": 0.9294752478599548, + "learning_rate": 3.668121010991115e-05, + "loss": 0.3099, + "num_input_tokens_seen": 4204304, + "step": 15670 + }, + { + "epoch": 4.108781127129751, + "grad_norm": 1.2306452989578247, + "learning_rate": 3.66710968884049e-05, + "loss": 0.3396, + "num_input_tokens_seen": 4205888, + "step": 15675 + }, + { + "epoch": 4.110091743119266, + "grad_norm": 0.5814481973648071, + "learning_rate": 3.666098122415823e-05, + "loss": 0.3335, + "num_input_tokens_seen": 4207728, + "step": 15680 + }, + { + "epoch": 4.111402359108781, + "grad_norm": 2.5774385929107666, + "learning_rate": 3.665086311928834e-05, + "loss": 0.358, + "num_input_tokens_seen": 4208848, + "step": 15685 + }, + { + "epoch": 4.1127129750982965, + "grad_norm": 0.9578647017478943, + "learning_rate": 3.664074257591293e-05, + "loss": 0.2864, + "num_input_tokens_seen": 4210144, + "step": 15690 + }, + { + "epoch": 4.114023591087811, + "grad_norm": 1.2046436071395874, + "learning_rate": 3.6630619596150225e-05, + "loss": 0.3717, + "num_input_tokens_seen": 4211488, + "step": 15695 + }, + { + "epoch": 4.115334207077327, + "grad_norm": 0.6489711403846741, + "learning_rate": 3.662049418211892e-05, + "loss": 0.3887, + "num_input_tokens_seen": 4213184, + "step": 15700 + }, + { + "epoch": 4.116644823066841, + "grad_norm": 3.897498846054077, + "learning_rate": 3.6610366335938266e-05, + "loss": 0.5167, + "num_input_tokens_seen": 4214624, + "step": 15705 + }, + { + "epoch": 4.117955439056357, + "grad_norm": 1.0137646198272705, + "learning_rate": 3.6600236059728e-05, + "loss": 0.3358, + "num_input_tokens_seen": 4215536, + "step": 15710 + }, + { + "epoch": 4.1192660550458715, + "grad_norm": 1.104446291923523, + "learning_rate": 3.6590103355608374e-05, + "loss": 0.4199, + "num_input_tokens_seen": 4216624, + "step": 15715 + }, + { + "epoch": 4.120576671035387, + "grad_norm": 1.544631838798523, + "learning_rate": 3.657996822570013e-05, + "loss": 0.4896, + "num_input_tokens_seen": 4218048, + "step": 15720 + }, + { + "epoch": 4.121887287024902, + "grad_norm": 0.619940459728241, + "learning_rate": 3.656983067212456e-05, + "loss": 0.3197, + "num_input_tokens_seen": 4219552, + "step": 15725 + }, + { + "epoch": 4.123197903014416, + "grad_norm": 1.1286085844039917, + "learning_rate": 3.655969069700341e-05, + "loss": 0.3096, + "num_input_tokens_seen": 4220528, + "step": 15730 + }, + { + "epoch": 4.124508519003932, + "grad_norm": 1.1023441553115845, + "learning_rate": 3.654954830245898e-05, + "loss": 0.3777, + "num_input_tokens_seen": 4221552, + "step": 15735 + }, + { + "epoch": 4.1258191349934465, + "grad_norm": 1.3243792057037354, + "learning_rate": 3.6539403490614046e-05, + "loss": 0.3656, + "num_input_tokens_seen": 4222784, + "step": 15740 + }, + { + "epoch": 4.127129750982962, + "grad_norm": 1.5740662813186646, + "learning_rate": 3.6529256263591885e-05, + "loss": 0.3299, + "num_input_tokens_seen": 4223840, + "step": 15745 + }, + { + "epoch": 4.128440366972477, + "grad_norm": 1.2841876745224, + "learning_rate": 3.651910662351632e-05, + "loss": 0.5596, + "num_input_tokens_seen": 4225728, + "step": 15750 + }, + { + "epoch": 4.129750982961992, + "grad_norm": 0.8785960078239441, + "learning_rate": 3.650895457251165e-05, + "loss": 0.3168, + "num_input_tokens_seen": 4227120, + "step": 15755 + }, + { + "epoch": 4.131061598951507, + "grad_norm": 5.025246620178223, + "learning_rate": 3.6498800112702664e-05, + "loss": 0.3855, + "num_input_tokens_seen": 4228240, + "step": 15760 + }, + { + "epoch": 4.1323722149410225, + "grad_norm": 1.9955869913101196, + "learning_rate": 3.648864324621468e-05, + "loss": 0.414, + "num_input_tokens_seen": 4229696, + "step": 15765 + }, + { + "epoch": 4.133682830930537, + "grad_norm": 0.8317142128944397, + "learning_rate": 3.647848397517353e-05, + "loss": 0.3102, + "num_input_tokens_seen": 4231024, + "step": 15770 + }, + { + "epoch": 4.134993446920053, + "grad_norm": 1.341521143913269, + "learning_rate": 3.6468322301705504e-05, + "loss": 0.3132, + "num_input_tokens_seen": 4232800, + "step": 15775 + }, + { + "epoch": 4.136304062909567, + "grad_norm": 1.3683067560195923, + "learning_rate": 3.6458158227937435e-05, + "loss": 0.3735, + "num_input_tokens_seen": 4234192, + "step": 15780 + }, + { + "epoch": 4.137614678899083, + "grad_norm": 0.9885950684547424, + "learning_rate": 3.6447991755996654e-05, + "loss": 0.3134, + "num_input_tokens_seen": 4235296, + "step": 15785 + }, + { + "epoch": 4.1389252948885975, + "grad_norm": 0.8239471316337585, + "learning_rate": 3.643782288801098e-05, + "loss": 0.4778, + "num_input_tokens_seen": 4236752, + "step": 15790 + }, + { + "epoch": 4.140235910878113, + "grad_norm": 0.9818436503410339, + "learning_rate": 3.642765162610873e-05, + "loss": 0.1812, + "num_input_tokens_seen": 4237856, + "step": 15795 + }, + { + "epoch": 4.141546526867628, + "grad_norm": 1.4910993576049805, + "learning_rate": 3.6417477972418745e-05, + "loss": 0.4212, + "num_input_tokens_seen": 4239280, + "step": 15800 + }, + { + "epoch": 4.142857142857143, + "grad_norm": 0.47369781136512756, + "learning_rate": 3.640730192907035e-05, + "loss": 0.4128, + "num_input_tokens_seen": 4240912, + "step": 15805 + }, + { + "epoch": 4.144167758846658, + "grad_norm": 1.1921430826187134, + "learning_rate": 3.639712349819337e-05, + "loss": 0.1849, + "num_input_tokens_seen": 4241808, + "step": 15810 + }, + { + "epoch": 4.145478374836173, + "grad_norm": 1.1198195219039917, + "learning_rate": 3.6386942681918135e-05, + "loss": 0.364, + "num_input_tokens_seen": 4242928, + "step": 15815 + }, + { + "epoch": 4.146788990825688, + "grad_norm": 0.8306604623794556, + "learning_rate": 3.637675948237547e-05, + "loss": 0.2223, + "num_input_tokens_seen": 4244048, + "step": 15820 + }, + { + "epoch": 4.148099606815203, + "grad_norm": 1.1642240285873413, + "learning_rate": 3.6366573901696704e-05, + "loss": 0.1887, + "num_input_tokens_seen": 4245104, + "step": 15825 + }, + { + "epoch": 4.149410222804718, + "grad_norm": 2.0896522998809814, + "learning_rate": 3.635638594201366e-05, + "loss": 0.385, + "num_input_tokens_seen": 4246192, + "step": 15830 + }, + { + "epoch": 4.150720838794233, + "grad_norm": 1.101740837097168, + "learning_rate": 3.6346195605458664e-05, + "loss": 0.299, + "num_input_tokens_seen": 4247312, + "step": 15835 + }, + { + "epoch": 4.1520314547837485, + "grad_norm": 1.4082634449005127, + "learning_rate": 3.633600289416453e-05, + "loss": 0.2227, + "num_input_tokens_seen": 4248368, + "step": 15840 + }, + { + "epoch": 4.153342070773263, + "grad_norm": 1.1146495342254639, + "learning_rate": 3.632580781026457e-05, + "loss": 0.3589, + "num_input_tokens_seen": 4249872, + "step": 15845 + }, + { + "epoch": 4.154652686762779, + "grad_norm": 0.9412559866905212, + "learning_rate": 3.631561035589262e-05, + "loss": 0.3052, + "num_input_tokens_seen": 4252336, + "step": 15850 + }, + { + "epoch": 4.155963302752293, + "grad_norm": 0.7068771123886108, + "learning_rate": 3.630541053318297e-05, + "loss": 0.2631, + "num_input_tokens_seen": 4253920, + "step": 15855 + }, + { + "epoch": 4.157273918741809, + "grad_norm": 1.062878131866455, + "learning_rate": 3.6295208344270415e-05, + "loss": 0.4785, + "num_input_tokens_seen": 4255312, + "step": 15860 + }, + { + "epoch": 4.1585845347313235, + "grad_norm": 1.4965165853500366, + "learning_rate": 3.628500379129028e-05, + "loss": 0.3564, + "num_input_tokens_seen": 4256400, + "step": 15865 + }, + { + "epoch": 4.159895150720839, + "grad_norm": 0.8750596642494202, + "learning_rate": 3.627479687637836e-05, + "loss": 0.2719, + "num_input_tokens_seen": 4258896, + "step": 15870 + }, + { + "epoch": 4.161205766710354, + "grad_norm": 1.6700485944747925, + "learning_rate": 3.6264587601670924e-05, + "loss": 0.3799, + "num_input_tokens_seen": 4260192, + "step": 15875 + }, + { + "epoch": 4.162516382699869, + "grad_norm": 0.9973132014274597, + "learning_rate": 3.625437596930476e-05, + "loss": 0.3969, + "num_input_tokens_seen": 4261280, + "step": 15880 + }, + { + "epoch": 4.163826998689384, + "grad_norm": 0.7257230281829834, + "learning_rate": 3.6244161981417165e-05, + "loss": 0.2885, + "num_input_tokens_seen": 4262192, + "step": 15885 + }, + { + "epoch": 4.165137614678899, + "grad_norm": 1.3630179166793823, + "learning_rate": 3.62339456401459e-05, + "loss": 0.3233, + "num_input_tokens_seen": 4263392, + "step": 15890 + }, + { + "epoch": 4.166448230668414, + "grad_norm": 7.083172798156738, + "learning_rate": 3.622372694762921e-05, + "loss": 0.5739, + "num_input_tokens_seen": 4264896, + "step": 15895 + }, + { + "epoch": 4.16775884665793, + "grad_norm": 1.7594884634017944, + "learning_rate": 3.621350590600587e-05, + "loss": 0.3867, + "num_input_tokens_seen": 4265984, + "step": 15900 + }, + { + "epoch": 4.169069462647444, + "grad_norm": 1.2484571933746338, + "learning_rate": 3.6203282517415114e-05, + "loss": 0.3354, + "num_input_tokens_seen": 4266864, + "step": 15905 + }, + { + "epoch": 4.17038007863696, + "grad_norm": 0.8981589078903198, + "learning_rate": 3.619305678399669e-05, + "loss": 0.3656, + "num_input_tokens_seen": 4268208, + "step": 15910 + }, + { + "epoch": 4.1716906946264745, + "grad_norm": 1.1001126766204834, + "learning_rate": 3.6182828707890816e-05, + "loss": 0.4248, + "num_input_tokens_seen": 4269392, + "step": 15915 + }, + { + "epoch": 4.173001310615989, + "grad_norm": 1.2208367586135864, + "learning_rate": 3.617259829123822e-05, + "loss": 0.475, + "num_input_tokens_seen": 4270800, + "step": 15920 + }, + { + "epoch": 4.174311926605505, + "grad_norm": 4.631033420562744, + "learning_rate": 3.6162365536180105e-05, + "loss": 0.4286, + "num_input_tokens_seen": 4271840, + "step": 15925 + }, + { + "epoch": 4.175622542595019, + "grad_norm": 0.8284432291984558, + "learning_rate": 3.615213044485817e-05, + "loss": 0.284, + "num_input_tokens_seen": 4273056, + "step": 15930 + }, + { + "epoch": 4.176933158584535, + "grad_norm": 1.738452434539795, + "learning_rate": 3.6141893019414606e-05, + "loss": 0.2853, + "num_input_tokens_seen": 4274112, + "step": 15935 + }, + { + "epoch": 4.1782437745740495, + "grad_norm": 3.140263319015503, + "learning_rate": 3.6131653261992085e-05, + "loss": 0.4569, + "num_input_tokens_seen": 4275344, + "step": 15940 + }, + { + "epoch": 4.179554390563565, + "grad_norm": 1.123582363128662, + "learning_rate": 3.612141117473377e-05, + "loss": 0.3603, + "num_input_tokens_seen": 4278688, + "step": 15945 + }, + { + "epoch": 4.18086500655308, + "grad_norm": 2.483088254928589, + "learning_rate": 3.61111667597833e-05, + "loss": 0.4292, + "num_input_tokens_seen": 4279920, + "step": 15950 + }, + { + "epoch": 4.182175622542595, + "grad_norm": 1.3998124599456787, + "learning_rate": 3.610092001928484e-05, + "loss": 0.4402, + "num_input_tokens_seen": 4280960, + "step": 15955 + }, + { + "epoch": 4.18348623853211, + "grad_norm": 0.8310244083404541, + "learning_rate": 3.6090670955383e-05, + "loss": 0.3392, + "num_input_tokens_seen": 4282416, + "step": 15960 + }, + { + "epoch": 4.184796854521625, + "grad_norm": 2.1219675540924072, + "learning_rate": 3.6080419570222885e-05, + "loss": 0.3891, + "num_input_tokens_seen": 4283584, + "step": 15965 + }, + { + "epoch": 4.18610747051114, + "grad_norm": 1.7568321228027344, + "learning_rate": 3.60701658659501e-05, + "loss": 0.3512, + "num_input_tokens_seen": 4284864, + "step": 15970 + }, + { + "epoch": 4.187418086500656, + "grad_norm": 0.9243115782737732, + "learning_rate": 3.605990984471073e-05, + "loss": 0.4118, + "num_input_tokens_seen": 4286320, + "step": 15975 + }, + { + "epoch": 4.18872870249017, + "grad_norm": 1.47775399684906, + "learning_rate": 3.6049651508651336e-05, + "loss": 0.2068, + "num_input_tokens_seen": 4287200, + "step": 15980 + }, + { + "epoch": 4.190039318479686, + "grad_norm": 1.367666244506836, + "learning_rate": 3.603939085991898e-05, + "loss": 0.3809, + "num_input_tokens_seen": 4288368, + "step": 15985 + }, + { + "epoch": 4.1913499344692005, + "grad_norm": 1.4481604099273682, + "learning_rate": 3.6029127900661196e-05, + "loss": 0.292, + "num_input_tokens_seen": 4289344, + "step": 15990 + }, + { + "epoch": 4.192660550458716, + "grad_norm": 0.9982228875160217, + "learning_rate": 3.601886263302599e-05, + "loss": 0.3997, + "num_input_tokens_seen": 4291568, + "step": 15995 + }, + { + "epoch": 4.193971166448231, + "grad_norm": 4.87637186050415, + "learning_rate": 3.600859505916187e-05, + "loss": 0.2934, + "num_input_tokens_seen": 4292688, + "step": 16000 + }, + { + "epoch": 4.195281782437746, + "grad_norm": 1.3762240409851074, + "learning_rate": 3.599832518121784e-05, + "loss": 0.3692, + "num_input_tokens_seen": 4294352, + "step": 16005 + }, + { + "epoch": 4.196592398427261, + "grad_norm": 4.966961860656738, + "learning_rate": 3.598805300134334e-05, + "loss": 0.4057, + "num_input_tokens_seen": 4295520, + "step": 16010 + }, + { + "epoch": 4.1979030144167755, + "grad_norm": 2.535316228866577, + "learning_rate": 3.597777852168834e-05, + "loss": 0.427, + "num_input_tokens_seen": 4296528, + "step": 16015 + }, + { + "epoch": 4.199213630406291, + "grad_norm": 1.4320895671844482, + "learning_rate": 3.596750174440326e-05, + "loss": 0.5832, + "num_input_tokens_seen": 4297824, + "step": 16020 + }, + { + "epoch": 4.200524246395806, + "grad_norm": 2.4130537509918213, + "learning_rate": 3.595722267163901e-05, + "loss": 0.2484, + "num_input_tokens_seen": 4299056, + "step": 16025 + }, + { + "epoch": 4.201834862385321, + "grad_norm": 1.3018770217895508, + "learning_rate": 3.5946941305547e-05, + "loss": 0.4867, + "num_input_tokens_seen": 4300400, + "step": 16030 + }, + { + "epoch": 4.203145478374836, + "grad_norm": 0.9917336702346802, + "learning_rate": 3.593665764827908e-05, + "loss": 0.4079, + "num_input_tokens_seen": 4301552, + "step": 16035 + }, + { + "epoch": 4.204456094364351, + "grad_norm": 0.8159061074256897, + "learning_rate": 3.592637170198762e-05, + "loss": 0.3235, + "num_input_tokens_seen": 4302736, + "step": 16040 + }, + { + "epoch": 4.205766710353866, + "grad_norm": 2.022294759750366, + "learning_rate": 3.591608346882543e-05, + "loss": 0.5273, + "num_input_tokens_seen": 4304048, + "step": 16045 + }, + { + "epoch": 4.207077326343382, + "grad_norm": 1.0471367835998535, + "learning_rate": 3.590579295094584e-05, + "loss": 0.5366, + "num_input_tokens_seen": 4305968, + "step": 16050 + }, + { + "epoch": 4.208387942332896, + "grad_norm": 0.9065966010093689, + "learning_rate": 3.589550015050263e-05, + "loss": 0.5183, + "num_input_tokens_seen": 4307216, + "step": 16055 + }, + { + "epoch": 4.209698558322412, + "grad_norm": 0.8917028307914734, + "learning_rate": 3.5885205069650056e-05, + "loss": 0.2713, + "num_input_tokens_seen": 4308336, + "step": 16060 + }, + { + "epoch": 4.2110091743119265, + "grad_norm": 0.8818591833114624, + "learning_rate": 3.587490771054288e-05, + "loss": 0.3436, + "num_input_tokens_seen": 4310016, + "step": 16065 + }, + { + "epoch": 4.212319790301442, + "grad_norm": 1.1182774305343628, + "learning_rate": 3.586460807533631e-05, + "loss": 0.8102, + "num_input_tokens_seen": 4311136, + "step": 16070 + }, + { + "epoch": 4.213630406290957, + "grad_norm": 7.3088250160217285, + "learning_rate": 3.585430616618604e-05, + "loss": 0.3118, + "num_input_tokens_seen": 4312384, + "step": 16075 + }, + { + "epoch": 4.214941022280472, + "grad_norm": 0.8767379522323608, + "learning_rate": 3.584400198524824e-05, + "loss": 0.3032, + "num_input_tokens_seen": 4314544, + "step": 16080 + }, + { + "epoch": 4.216251638269987, + "grad_norm": 1.0653815269470215, + "learning_rate": 3.583369553467956e-05, + "loss": 0.294, + "num_input_tokens_seen": 4316176, + "step": 16085 + }, + { + "epoch": 4.217562254259502, + "grad_norm": 0.4377278685569763, + "learning_rate": 3.582338681663713e-05, + "loss": 0.2531, + "num_input_tokens_seen": 4317312, + "step": 16090 + }, + { + "epoch": 4.218872870249017, + "grad_norm": 1.200290322303772, + "learning_rate": 3.581307583327854e-05, + "loss": 0.5061, + "num_input_tokens_seen": 4318768, + "step": 16095 + }, + { + "epoch": 4.220183486238533, + "grad_norm": 0.7828378677368164, + "learning_rate": 3.580276258676186e-05, + "loss": 0.4635, + "num_input_tokens_seen": 4319968, + "step": 16100 + }, + { + "epoch": 4.221494102228047, + "grad_norm": 0.8747138381004333, + "learning_rate": 3.579244707924564e-05, + "loss": 0.4661, + "num_input_tokens_seen": 4321072, + "step": 16105 + }, + { + "epoch": 4.222804718217562, + "grad_norm": 0.8140920400619507, + "learning_rate": 3.5782129312888885e-05, + "loss": 0.2692, + "num_input_tokens_seen": 4322192, + "step": 16110 + }, + { + "epoch": 4.224115334207077, + "grad_norm": 2.6387248039245605, + "learning_rate": 3.5771809289851097e-05, + "loss": 0.4254, + "num_input_tokens_seen": 4323376, + "step": 16115 + }, + { + "epoch": 4.225425950196592, + "grad_norm": 0.612833559513092, + "learning_rate": 3.576148701229224e-05, + "loss": 0.3859, + "num_input_tokens_seen": 4324656, + "step": 16120 + }, + { + "epoch": 4.226736566186108, + "grad_norm": 1.3577735424041748, + "learning_rate": 3.575116248237275e-05, + "loss": 0.3459, + "num_input_tokens_seen": 4326208, + "step": 16125 + }, + { + "epoch": 4.228047182175622, + "grad_norm": 1.1258851289749146, + "learning_rate": 3.574083570225352e-05, + "loss": 0.3416, + "num_input_tokens_seen": 4327392, + "step": 16130 + }, + { + "epoch": 4.229357798165138, + "grad_norm": 1.365039587020874, + "learning_rate": 3.573050667409594e-05, + "loss": 0.343, + "num_input_tokens_seen": 4329024, + "step": 16135 + }, + { + "epoch": 4.2306684141546524, + "grad_norm": 1.292646884918213, + "learning_rate": 3.572017540006186e-05, + "loss": 0.3892, + "num_input_tokens_seen": 4330352, + "step": 16140 + }, + { + "epoch": 4.231979030144168, + "grad_norm": 1.5709598064422607, + "learning_rate": 3.570984188231359e-05, + "loss": 0.2301, + "num_input_tokens_seen": 4331488, + "step": 16145 + }, + { + "epoch": 4.233289646133683, + "grad_norm": 0.5847214460372925, + "learning_rate": 3.5699506123013906e-05, + "loss": 0.3914, + "num_input_tokens_seen": 4332608, + "step": 16150 + }, + { + "epoch": 4.234600262123198, + "grad_norm": 1.411670446395874, + "learning_rate": 3.568916812432609e-05, + "loss": 0.2776, + "num_input_tokens_seen": 4334208, + "step": 16155 + }, + { + "epoch": 4.235910878112713, + "grad_norm": 0.8293511867523193, + "learning_rate": 3.5678827888413856e-05, + "loss": 0.3225, + "num_input_tokens_seen": 4335424, + "step": 16160 + }, + { + "epoch": 4.237221494102228, + "grad_norm": 2.5333197116851807, + "learning_rate": 3.56684854174414e-05, + "loss": 0.3052, + "num_input_tokens_seen": 4336512, + "step": 16165 + }, + { + "epoch": 4.238532110091743, + "grad_norm": 0.16068056225776672, + "learning_rate": 3.5658140713573365e-05, + "loss": 0.1543, + "num_input_tokens_seen": 4337920, + "step": 16170 + }, + { + "epoch": 4.239842726081259, + "grad_norm": 1.1449573040008545, + "learning_rate": 3.56477937789749e-05, + "loss": 0.4791, + "num_input_tokens_seen": 4339264, + "step": 16175 + }, + { + "epoch": 4.241153342070773, + "grad_norm": 0.6628363132476807, + "learning_rate": 3.56374446158116e-05, + "loss": 0.2373, + "num_input_tokens_seen": 4340560, + "step": 16180 + }, + { + "epoch": 4.242463958060289, + "grad_norm": 0.7908562421798706, + "learning_rate": 3.5627093226249524e-05, + "loss": 0.2482, + "num_input_tokens_seen": 4342000, + "step": 16185 + }, + { + "epoch": 4.243774574049803, + "grad_norm": 1.1274621486663818, + "learning_rate": 3.5616739612455196e-05, + "loss": 0.4551, + "num_input_tokens_seen": 4343568, + "step": 16190 + }, + { + "epoch": 4.245085190039318, + "grad_norm": 0.6192523837089539, + "learning_rate": 3.560638377659561e-05, + "loss": 0.3042, + "num_input_tokens_seen": 4345136, + "step": 16195 + }, + { + "epoch": 4.246395806028834, + "grad_norm": 1.6126611232757568, + "learning_rate": 3.559602572083822e-05, + "loss": 0.2995, + "num_input_tokens_seen": 4346240, + "step": 16200 + }, + { + "epoch": 4.247706422018348, + "grad_norm": 0.7478029131889343, + "learning_rate": 3.558566544735096e-05, + "loss": 0.3699, + "num_input_tokens_seen": 4347728, + "step": 16205 + }, + { + "epoch": 4.249017038007864, + "grad_norm": 0.9968231320381165, + "learning_rate": 3.5575302958302214e-05, + "loss": 0.4468, + "num_input_tokens_seen": 4349168, + "step": 16210 + }, + { + "epoch": 4.250327653997378, + "grad_norm": 1.8321645259857178, + "learning_rate": 3.556493825586083e-05, + "loss": 0.2517, + "num_input_tokens_seen": 4350432, + "step": 16215 + }, + { + "epoch": 4.251638269986894, + "grad_norm": 1.5205355882644653, + "learning_rate": 3.555457134219613e-05, + "loss": 0.4911, + "num_input_tokens_seen": 4351600, + "step": 16220 + }, + { + "epoch": 4.252948885976409, + "grad_norm": 1.257577896118164, + "learning_rate": 3.554420221947788e-05, + "loss": 0.6904, + "num_input_tokens_seen": 4352784, + "step": 16225 + }, + { + "epoch": 4.254259501965924, + "grad_norm": 2.587339401245117, + "learning_rate": 3.553383088987632e-05, + "loss": 0.5542, + "num_input_tokens_seen": 4353952, + "step": 16230 + }, + { + "epoch": 4.255570117955439, + "grad_norm": 2.0646708011627197, + "learning_rate": 3.552345735556216e-05, + "loss": 0.3483, + "num_input_tokens_seen": 4355328, + "step": 16235 + }, + { + "epoch": 4.256880733944954, + "grad_norm": 1.1129705905914307, + "learning_rate": 3.5513081618706555e-05, + "loss": 0.3568, + "num_input_tokens_seen": 4356672, + "step": 16240 + }, + { + "epoch": 4.258191349934469, + "grad_norm": 1.2167717218399048, + "learning_rate": 3.550270368148113e-05, + "loss": 0.3782, + "num_input_tokens_seen": 4357840, + "step": 16245 + }, + { + "epoch": 4.259501965923985, + "grad_norm": 0.8915961384773254, + "learning_rate": 3.549232354605798e-05, + "loss": 0.3746, + "num_input_tokens_seen": 4358816, + "step": 16250 + }, + { + "epoch": 4.260812581913499, + "grad_norm": 1.9191224575042725, + "learning_rate": 3.548194121460964e-05, + "loss": 0.3762, + "num_input_tokens_seen": 4359920, + "step": 16255 + }, + { + "epoch": 4.262123197903015, + "grad_norm": 2.3277933597564697, + "learning_rate": 3.547155668930912e-05, + "loss": 0.3764, + "num_input_tokens_seen": 4360944, + "step": 16260 + }, + { + "epoch": 4.263433813892529, + "grad_norm": 0.969647228717804, + "learning_rate": 3.546116997232987e-05, + "loss": 0.2993, + "num_input_tokens_seen": 4362112, + "step": 16265 + }, + { + "epoch": 4.264744429882045, + "grad_norm": 0.7612922191619873, + "learning_rate": 3.545078106584582e-05, + "loss": 0.3334, + "num_input_tokens_seen": 4363232, + "step": 16270 + }, + { + "epoch": 4.26605504587156, + "grad_norm": 0.6265087723731995, + "learning_rate": 3.544038997203135e-05, + "loss": 0.2426, + "num_input_tokens_seen": 4364800, + "step": 16275 + }, + { + "epoch": 4.267365661861074, + "grad_norm": 1.6366422176361084, + "learning_rate": 3.5429996693061304e-05, + "loss": 0.4962, + "num_input_tokens_seen": 4366096, + "step": 16280 + }, + { + "epoch": 4.26867627785059, + "grad_norm": 1.1425957679748535, + "learning_rate": 3.541960123111097e-05, + "loss": 0.2441, + "num_input_tokens_seen": 4367744, + "step": 16285 + }, + { + "epoch": 4.269986893840104, + "grad_norm": 1.2478276491165161, + "learning_rate": 3.54092035883561e-05, + "loss": 0.3248, + "num_input_tokens_seen": 4369040, + "step": 16290 + }, + { + "epoch": 4.27129750982962, + "grad_norm": 3.391400098800659, + "learning_rate": 3.53988037669729e-05, + "loss": 0.5063, + "num_input_tokens_seen": 4370784, + "step": 16295 + }, + { + "epoch": 4.272608125819135, + "grad_norm": 0.9003692269325256, + "learning_rate": 3.5388401769138053e-05, + "loss": 0.2924, + "num_input_tokens_seen": 4372880, + "step": 16300 + }, + { + "epoch": 4.27391874180865, + "grad_norm": 1.889723777770996, + "learning_rate": 3.5377997597028656e-05, + "loss": 0.2585, + "num_input_tokens_seen": 4373840, + "step": 16305 + }, + { + "epoch": 4.275229357798165, + "grad_norm": 0.8658744692802429, + "learning_rate": 3.53675912528223e-05, + "loss": 0.3419, + "num_input_tokens_seen": 4375536, + "step": 16310 + }, + { + "epoch": 4.27653997378768, + "grad_norm": 1.1918816566467285, + "learning_rate": 3.535718273869699e-05, + "loss": 0.2677, + "num_input_tokens_seen": 4376464, + "step": 16315 + }, + { + "epoch": 4.277850589777195, + "grad_norm": 1.3341988325119019, + "learning_rate": 3.534677205683125e-05, + "loss": 0.3415, + "num_input_tokens_seen": 4377776, + "step": 16320 + }, + { + "epoch": 4.2791612057667106, + "grad_norm": 1.1043376922607422, + "learning_rate": 3.533635920940399e-05, + "loss": 0.4136, + "num_input_tokens_seen": 4379120, + "step": 16325 + }, + { + "epoch": 4.280471821756225, + "grad_norm": 0.6991418600082397, + "learning_rate": 3.5325944198594607e-05, + "loss": 0.3219, + "num_input_tokens_seen": 4380352, + "step": 16330 + }, + { + "epoch": 4.281782437745741, + "grad_norm": 0.7873498201370239, + "learning_rate": 3.531552702658294e-05, + "loss": 0.2588, + "num_input_tokens_seen": 4381488, + "step": 16335 + }, + { + "epoch": 4.283093053735255, + "grad_norm": 1.1278334856033325, + "learning_rate": 3.53051076955493e-05, + "loss": 0.3056, + "num_input_tokens_seen": 4382672, + "step": 16340 + }, + { + "epoch": 4.284403669724771, + "grad_norm": 2.03818678855896, + "learning_rate": 3.5294686207674414e-05, + "loss": 0.4884, + "num_input_tokens_seen": 4384000, + "step": 16345 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 1.5811673402786255, + "learning_rate": 3.5284262565139494e-05, + "loss": 0.3466, + "num_input_tokens_seen": 4385168, + "step": 16350 + }, + { + "epoch": 4.287024901703801, + "grad_norm": 1.6101713180541992, + "learning_rate": 3.5273836770126204e-05, + "loss": 0.3894, + "num_input_tokens_seen": 4386400, + "step": 16355 + }, + { + "epoch": 4.288335517693316, + "grad_norm": 0.5485197901725769, + "learning_rate": 3.5263408824816615e-05, + "loss": 0.3421, + "num_input_tokens_seen": 4387840, + "step": 16360 + }, + { + "epoch": 4.289646133682831, + "grad_norm": 0.5629456043243408, + "learning_rate": 3.52529787313933e-05, + "loss": 0.4295, + "num_input_tokens_seen": 4389440, + "step": 16365 + }, + { + "epoch": 4.290956749672346, + "grad_norm": 1.3082325458526611, + "learning_rate": 3.524254649203925e-05, + "loss": 0.2494, + "num_input_tokens_seen": 4390848, + "step": 16370 + }, + { + "epoch": 4.292267365661861, + "grad_norm": 0.9820791482925415, + "learning_rate": 3.5232112108937914e-05, + "loss": 0.3937, + "num_input_tokens_seen": 4392208, + "step": 16375 + }, + { + "epoch": 4.293577981651376, + "grad_norm": 0.963915228843689, + "learning_rate": 3.52216755842732e-05, + "loss": 0.5803, + "num_input_tokens_seen": 4393968, + "step": 16380 + }, + { + "epoch": 4.294888597640891, + "grad_norm": 1.561095952987671, + "learning_rate": 3.521123692022944e-05, + "loss": 0.4407, + "num_input_tokens_seen": 4394944, + "step": 16385 + }, + { + "epoch": 4.296199213630406, + "grad_norm": 1.2331210374832153, + "learning_rate": 3.520079611899144e-05, + "loss": 0.328, + "num_input_tokens_seen": 4396112, + "step": 16390 + }, + { + "epoch": 4.297509829619921, + "grad_norm": 0.7122417688369751, + "learning_rate": 3.519035318274444e-05, + "loss": 0.3313, + "num_input_tokens_seen": 4397376, + "step": 16395 + }, + { + "epoch": 4.2988204456094365, + "grad_norm": 1.878895878791809, + "learning_rate": 3.517990811367412e-05, + "loss": 0.4583, + "num_input_tokens_seen": 4398624, + "step": 16400 + }, + { + "epoch": 4.300131061598951, + "grad_norm": 2.4921514987945557, + "learning_rate": 3.516946091396662e-05, + "loss": 0.4124, + "num_input_tokens_seen": 4399744, + "step": 16405 + }, + { + "epoch": 4.301441677588467, + "grad_norm": 6.560815811157227, + "learning_rate": 3.5159011585808525e-05, + "loss": 0.5516, + "num_input_tokens_seen": 4400720, + "step": 16410 + }, + { + "epoch": 4.302752293577981, + "grad_norm": 1.3412381410598755, + "learning_rate": 3.5148560131386867e-05, + "loss": 0.4549, + "num_input_tokens_seen": 4402464, + "step": 16415 + }, + { + "epoch": 4.304062909567497, + "grad_norm": 0.6618605852127075, + "learning_rate": 3.5138106552889106e-05, + "loss": 0.3271, + "num_input_tokens_seen": 4403968, + "step": 16420 + }, + { + "epoch": 4.305373525557012, + "grad_norm": 1.4042762517929077, + "learning_rate": 3.5127650852503156e-05, + "loss": 0.233, + "num_input_tokens_seen": 4405200, + "step": 16425 + }, + { + "epoch": 4.306684141546527, + "grad_norm": 2.3402645587921143, + "learning_rate": 3.511719303241738e-05, + "loss": 0.2064, + "num_input_tokens_seen": 4406432, + "step": 16430 + }, + { + "epoch": 4.307994757536042, + "grad_norm": 3.627373456954956, + "learning_rate": 3.5106733094820586e-05, + "loss": 0.4822, + "num_input_tokens_seen": 4407648, + "step": 16435 + }, + { + "epoch": 4.309305373525557, + "grad_norm": 1.179955244064331, + "learning_rate": 3.509627104190202e-05, + "loss": 0.3385, + "num_input_tokens_seen": 4408736, + "step": 16440 + }, + { + "epoch": 4.310615989515072, + "grad_norm": 2.536435127258301, + "learning_rate": 3.5085806875851365e-05, + "loss": 0.3794, + "num_input_tokens_seen": 4409984, + "step": 16445 + }, + { + "epoch": 4.3119266055045875, + "grad_norm": 1.9880539178848267, + "learning_rate": 3.507534059885876e-05, + "loss": 0.3326, + "num_input_tokens_seen": 4410960, + "step": 16450 + }, + { + "epoch": 4.313237221494102, + "grad_norm": 1.6764252185821533, + "learning_rate": 3.5064872213114784e-05, + "loss": 0.413, + "num_input_tokens_seen": 4412016, + "step": 16455 + }, + { + "epoch": 4.314547837483618, + "grad_norm": 0.7352668642997742, + "learning_rate": 3.505440172081044e-05, + "loss": 0.3325, + "num_input_tokens_seen": 4413776, + "step": 16460 + }, + { + "epoch": 4.315858453473132, + "grad_norm": 1.3017317056655884, + "learning_rate": 3.504392912413719e-05, + "loss": 0.2539, + "num_input_tokens_seen": 4414720, + "step": 16465 + }, + { + "epoch": 4.317169069462647, + "grad_norm": 1.069636344909668, + "learning_rate": 3.503345442528692e-05, + "loss": 0.341, + "num_input_tokens_seen": 4415824, + "step": 16470 + }, + { + "epoch": 4.3184796854521625, + "grad_norm": 1.3248212337493896, + "learning_rate": 3.5022977626451985e-05, + "loss": 0.2146, + "num_input_tokens_seen": 4416832, + "step": 16475 + }, + { + "epoch": 4.319790301441677, + "grad_norm": 1.3120999336242676, + "learning_rate": 3.501249872982515e-05, + "loss": 0.4066, + "num_input_tokens_seen": 4418176, + "step": 16480 + }, + { + "epoch": 4.321100917431193, + "grad_norm": 0.9720901846885681, + "learning_rate": 3.500201773759963e-05, + "loss": 0.382, + "num_input_tokens_seen": 4419504, + "step": 16485 + }, + { + "epoch": 4.322411533420707, + "grad_norm": 2.136563777923584, + "learning_rate": 3.499153465196907e-05, + "loss": 0.6246, + "num_input_tokens_seen": 4420784, + "step": 16490 + }, + { + "epoch": 4.323722149410223, + "grad_norm": 0.923124372959137, + "learning_rate": 3.4981049475127584e-05, + "loss": 0.4408, + "num_input_tokens_seen": 4421968, + "step": 16495 + }, + { + "epoch": 4.325032765399738, + "grad_norm": 0.6953665614128113, + "learning_rate": 3.4970562209269684e-05, + "loss": 0.2405, + "num_input_tokens_seen": 4423088, + "step": 16500 + }, + { + "epoch": 4.326343381389253, + "grad_norm": 1.0512827634811401, + "learning_rate": 3.496007285659034e-05, + "loss": 0.5313, + "num_input_tokens_seen": 4424672, + "step": 16505 + }, + { + "epoch": 4.327653997378768, + "grad_norm": 0.9727792143821716, + "learning_rate": 3.494958141928496e-05, + "loss": 0.2374, + "num_input_tokens_seen": 4425632, + "step": 16510 + }, + { + "epoch": 4.328964613368283, + "grad_norm": 0.7677636742591858, + "learning_rate": 3.4939087899549386e-05, + "loss": 0.3351, + "num_input_tokens_seen": 4426960, + "step": 16515 + }, + { + "epoch": 4.330275229357798, + "grad_norm": 1.2978479862213135, + "learning_rate": 3.4928592299579874e-05, + "loss": 0.3295, + "num_input_tokens_seen": 4428816, + "step": 16520 + }, + { + "epoch": 4.3315858453473135, + "grad_norm": 1.336276888847351, + "learning_rate": 3.4918094621573164e-05, + "loss": 0.3757, + "num_input_tokens_seen": 4429888, + "step": 16525 + }, + { + "epoch": 4.332896461336828, + "grad_norm": 0.6734774708747864, + "learning_rate": 3.490759486772637e-05, + "loss": 0.3723, + "num_input_tokens_seen": 4431424, + "step": 16530 + }, + { + "epoch": 4.334207077326344, + "grad_norm": 1.0017589330673218, + "learning_rate": 3.48970930402371e-05, + "loss": 0.2813, + "num_input_tokens_seen": 4432880, + "step": 16535 + }, + { + "epoch": 4.335517693315858, + "grad_norm": 0.8977063298225403, + "learning_rate": 3.488658914130336e-05, + "loss": 0.2878, + "num_input_tokens_seen": 4433920, + "step": 16540 + }, + { + "epoch": 4.336828309305374, + "grad_norm": 1.2458492517471313, + "learning_rate": 3.487608317312358e-05, + "loss": 0.3528, + "num_input_tokens_seen": 4435104, + "step": 16545 + }, + { + "epoch": 4.3381389252948885, + "grad_norm": 1.6608432531356812, + "learning_rate": 3.486557513789668e-05, + "loss": 0.3744, + "num_input_tokens_seen": 4436416, + "step": 16550 + }, + { + "epoch": 4.339449541284404, + "grad_norm": 1.3819411993026733, + "learning_rate": 3.485506503782194e-05, + "loss": 0.3684, + "num_input_tokens_seen": 4437600, + "step": 16555 + }, + { + "epoch": 4.340760157273919, + "grad_norm": 0.5283941626548767, + "learning_rate": 3.484455287509911e-05, + "loss": 0.3311, + "num_input_tokens_seen": 4438928, + "step": 16560 + }, + { + "epoch": 4.342070773263433, + "grad_norm": 0.7612589001655579, + "learning_rate": 3.483403865192838e-05, + "loss": 0.4488, + "num_input_tokens_seen": 4440400, + "step": 16565 + }, + { + "epoch": 4.343381389252949, + "grad_norm": 1.573409914970398, + "learning_rate": 3.482352237051035e-05, + "loss": 0.4396, + "num_input_tokens_seen": 4441728, + "step": 16570 + }, + { + "epoch": 4.344692005242464, + "grad_norm": 1.6060025691986084, + "learning_rate": 3.481300403304606e-05, + "loss": 0.3744, + "num_input_tokens_seen": 4442768, + "step": 16575 + }, + { + "epoch": 4.346002621231979, + "grad_norm": 0.9340465068817139, + "learning_rate": 3.480248364173698e-05, + "loss": 0.3073, + "num_input_tokens_seen": 4444448, + "step": 16580 + }, + { + "epoch": 4.347313237221494, + "grad_norm": 0.9318135976791382, + "learning_rate": 3.4791961198785014e-05, + "loss": 0.261, + "num_input_tokens_seen": 4445568, + "step": 16585 + }, + { + "epoch": 4.348623853211009, + "grad_norm": 2.047058582305908, + "learning_rate": 3.478143670639248e-05, + "loss": 0.3954, + "num_input_tokens_seen": 4446896, + "step": 16590 + }, + { + "epoch": 4.349934469200524, + "grad_norm": 0.9964312314987183, + "learning_rate": 3.477091016676216e-05, + "loss": 0.3284, + "num_input_tokens_seen": 4448192, + "step": 16595 + }, + { + "epoch": 4.3512450851900395, + "grad_norm": 0.6738991737365723, + "learning_rate": 3.476038158209722e-05, + "loss": 0.3918, + "num_input_tokens_seen": 4449296, + "step": 16600 + }, + { + "epoch": 4.352555701179554, + "grad_norm": 1.57590651512146, + "learning_rate": 3.4749850954601274e-05, + "loss": 0.3135, + "num_input_tokens_seen": 4451136, + "step": 16605 + }, + { + "epoch": 4.35386631716907, + "grad_norm": 0.6162951588630676, + "learning_rate": 3.473931828647837e-05, + "loss": 0.2657, + "num_input_tokens_seen": 4452480, + "step": 16610 + }, + { + "epoch": 4.355176933158584, + "grad_norm": 1.3472906351089478, + "learning_rate": 3.472878357993298e-05, + "loss": 0.3522, + "num_input_tokens_seen": 4453536, + "step": 16615 + }, + { + "epoch": 4.3564875491481, + "grad_norm": 1.0864195823669434, + "learning_rate": 3.4718246837169996e-05, + "loss": 0.5574, + "num_input_tokens_seen": 4455712, + "step": 16620 + }, + { + "epoch": 4.3577981651376145, + "grad_norm": 0.9263206124305725, + "learning_rate": 3.470770806039474e-05, + "loss": 0.3752, + "num_input_tokens_seen": 4457056, + "step": 16625 + }, + { + "epoch": 4.35910878112713, + "grad_norm": 1.175083041191101, + "learning_rate": 3.469716725181296e-05, + "loss": 0.3317, + "num_input_tokens_seen": 4458352, + "step": 16630 + }, + { + "epoch": 4.360419397116645, + "grad_norm": 2.8541009426116943, + "learning_rate": 3.468662441363083e-05, + "loss": 0.3502, + "num_input_tokens_seen": 4459456, + "step": 16635 + }, + { + "epoch": 4.36173001310616, + "grad_norm": 0.9156342148780823, + "learning_rate": 3.467607954805495e-05, + "loss": 0.5103, + "num_input_tokens_seen": 4460704, + "step": 16640 + }, + { + "epoch": 4.363040629095675, + "grad_norm": 0.9549800157546997, + "learning_rate": 3.466553265729233e-05, + "loss": 0.2875, + "num_input_tokens_seen": 4462192, + "step": 16645 + }, + { + "epoch": 4.3643512450851905, + "grad_norm": 1.9537122249603271, + "learning_rate": 3.465498374355044e-05, + "loss": 0.4002, + "num_input_tokens_seen": 4463328, + "step": 16650 + }, + { + "epoch": 4.365661861074705, + "grad_norm": 1.6562749147415161, + "learning_rate": 3.4644432809037134e-05, + "loss": 0.2635, + "num_input_tokens_seen": 4464384, + "step": 16655 + }, + { + "epoch": 4.36697247706422, + "grad_norm": 2.246049642562866, + "learning_rate": 3.463387985596069e-05, + "loss": 0.4116, + "num_input_tokens_seen": 4465808, + "step": 16660 + }, + { + "epoch": 4.368283093053735, + "grad_norm": 1.1102821826934814, + "learning_rate": 3.462332488652985e-05, + "loss": 0.421, + "num_input_tokens_seen": 4467200, + "step": 16665 + }, + { + "epoch": 4.36959370904325, + "grad_norm": 0.8886039853096008, + "learning_rate": 3.461276790295374e-05, + "loss": 0.3111, + "num_input_tokens_seen": 4468608, + "step": 16670 + }, + { + "epoch": 4.3709043250327655, + "grad_norm": 1.103330135345459, + "learning_rate": 3.4602208907441915e-05, + "loss": 0.2945, + "num_input_tokens_seen": 4469920, + "step": 16675 + }, + { + "epoch": 4.37221494102228, + "grad_norm": 1.0301135778427124, + "learning_rate": 3.4591647902204355e-05, + "loss": 0.4997, + "num_input_tokens_seen": 4470896, + "step": 16680 + }, + { + "epoch": 4.373525557011796, + "grad_norm": 0.8861894607543945, + "learning_rate": 3.4581084889451466e-05, + "loss": 0.4688, + "num_input_tokens_seen": 4472864, + "step": 16685 + }, + { + "epoch": 4.37483617300131, + "grad_norm": 1.0357921123504639, + "learning_rate": 3.457051987139406e-05, + "loss": 0.3674, + "num_input_tokens_seen": 4474032, + "step": 16690 + }, + { + "epoch": 4.376146788990826, + "grad_norm": 2.901556968688965, + "learning_rate": 3.455995285024338e-05, + "loss": 0.4606, + "num_input_tokens_seen": 4475728, + "step": 16695 + }, + { + "epoch": 4.3774574049803405, + "grad_norm": 1.0546369552612305, + "learning_rate": 3.454938382821109e-05, + "loss": 0.5578, + "num_input_tokens_seen": 4477344, + "step": 16700 + }, + { + "epoch": 4.378768020969856, + "grad_norm": 1.0054057836532593, + "learning_rate": 3.453881280750927e-05, + "loss": 0.3505, + "num_input_tokens_seen": 4478480, + "step": 16705 + }, + { + "epoch": 4.380078636959371, + "grad_norm": 1.2647837400436401, + "learning_rate": 3.452823979035041e-05, + "loss": 0.3172, + "num_input_tokens_seen": 4479552, + "step": 16710 + }, + { + "epoch": 4.381389252948886, + "grad_norm": 1.3683818578720093, + "learning_rate": 3.451766477894742e-05, + "loss": 0.4082, + "num_input_tokens_seen": 4480704, + "step": 16715 + }, + { + "epoch": 4.382699868938401, + "grad_norm": 2.102212905883789, + "learning_rate": 3.450708777551364e-05, + "loss": 0.4096, + "num_input_tokens_seen": 4482176, + "step": 16720 + }, + { + "epoch": 4.3840104849279165, + "grad_norm": 0.5716325044631958, + "learning_rate": 3.4496508782262804e-05, + "loss": 0.3224, + "num_input_tokens_seen": 4483376, + "step": 16725 + }, + { + "epoch": 4.385321100917431, + "grad_norm": 2.446758985519409, + "learning_rate": 3.4485927801409105e-05, + "loss": 0.3459, + "num_input_tokens_seen": 4484656, + "step": 16730 + }, + { + "epoch": 4.386631716906947, + "grad_norm": 0.761454701423645, + "learning_rate": 3.44753448351671e-05, + "loss": 0.3204, + "num_input_tokens_seen": 4485968, + "step": 16735 + }, + { + "epoch": 4.387942332896461, + "grad_norm": 1.8965692520141602, + "learning_rate": 3.4464759885751796e-05, + "loss": 0.3591, + "num_input_tokens_seen": 4487888, + "step": 16740 + }, + { + "epoch": 4.389252948885977, + "grad_norm": 0.64609295129776, + "learning_rate": 3.4454172955378595e-05, + "loss": 0.3396, + "num_input_tokens_seen": 4488944, + "step": 16745 + }, + { + "epoch": 4.3905635648754915, + "grad_norm": 0.8838377594947815, + "learning_rate": 3.444358404626335e-05, + "loss": 0.3313, + "num_input_tokens_seen": 4490336, + "step": 16750 + }, + { + "epoch": 4.391874180865006, + "grad_norm": 0.5723849534988403, + "learning_rate": 3.4432993160622256e-05, + "loss": 0.2999, + "num_input_tokens_seen": 4491504, + "step": 16755 + }, + { + "epoch": 4.393184796854522, + "grad_norm": 0.8249144554138184, + "learning_rate": 3.4422400300672e-05, + "loss": 0.4308, + "num_input_tokens_seen": 4492976, + "step": 16760 + }, + { + "epoch": 4.394495412844036, + "grad_norm": 1.4695228338241577, + "learning_rate": 3.441180546862965e-05, + "loss": 0.4989, + "num_input_tokens_seen": 4493968, + "step": 16765 + }, + { + "epoch": 4.395806028833552, + "grad_norm": 1.4284465312957764, + "learning_rate": 3.440120866671268e-05, + "loss": 0.5287, + "num_input_tokens_seen": 4495168, + "step": 16770 + }, + { + "epoch": 4.3971166448230665, + "grad_norm": 1.8091304302215576, + "learning_rate": 3.4390609897138965e-05, + "loss": 0.3499, + "num_input_tokens_seen": 4496256, + "step": 16775 + }, + { + "epoch": 4.398427260812582, + "grad_norm": 0.5677053332328796, + "learning_rate": 3.438000916212683e-05, + "loss": 0.2674, + "num_input_tokens_seen": 4497216, + "step": 16780 + }, + { + "epoch": 4.399737876802097, + "grad_norm": 1.3625829219818115, + "learning_rate": 3.436940646389499e-05, + "loss": 0.2457, + "num_input_tokens_seen": 4498448, + "step": 16785 + }, + { + "epoch": 4.401048492791612, + "grad_norm": 1.2813143730163574, + "learning_rate": 3.435880180466257e-05, + "loss": 0.3353, + "num_input_tokens_seen": 4499776, + "step": 16790 + }, + { + "epoch": 4.402359108781127, + "grad_norm": 1.2854602336883545, + "learning_rate": 3.434819518664909e-05, + "loss": 0.3804, + "num_input_tokens_seen": 4501104, + "step": 16795 + }, + { + "epoch": 4.4036697247706424, + "grad_norm": 0.8368495106697083, + "learning_rate": 3.4337586612074515e-05, + "loss": 0.3949, + "num_input_tokens_seen": 4504208, + "step": 16800 + }, + { + "epoch": 4.404980340760157, + "grad_norm": 1.118207573890686, + "learning_rate": 3.4326976083159194e-05, + "loss": 0.3852, + "num_input_tokens_seen": 4505920, + "step": 16805 + }, + { + "epoch": 4.406290956749673, + "grad_norm": 1.4319770336151123, + "learning_rate": 3.43163636021239e-05, + "loss": 0.4092, + "num_input_tokens_seen": 4507536, + "step": 16810 + }, + { + "epoch": 4.407601572739187, + "grad_norm": 2.1670684814453125, + "learning_rate": 3.43057491711898e-05, + "loss": 0.3021, + "num_input_tokens_seen": 4508768, + "step": 16815 + }, + { + "epoch": 4.408912188728703, + "grad_norm": 0.5807999968528748, + "learning_rate": 3.429513279257848e-05, + "loss": 0.3312, + "num_input_tokens_seen": 4510336, + "step": 16820 + }, + { + "epoch": 4.4102228047182175, + "grad_norm": 1.2359532117843628, + "learning_rate": 3.428451446851192e-05, + "loss": 0.36, + "num_input_tokens_seen": 4511840, + "step": 16825 + }, + { + "epoch": 4.411533420707733, + "grad_norm": 1.2827632427215576, + "learning_rate": 3.427389420121253e-05, + "loss": 0.3445, + "num_input_tokens_seen": 4513008, + "step": 16830 + }, + { + "epoch": 4.412844036697248, + "grad_norm": 1.095291018486023, + "learning_rate": 3.42632719929031e-05, + "loss": 0.2621, + "num_input_tokens_seen": 4514352, + "step": 16835 + }, + { + "epoch": 4.414154652686763, + "grad_norm": 1.087952733039856, + "learning_rate": 3.425264784580686e-05, + "loss": 0.3844, + "num_input_tokens_seen": 4515792, + "step": 16840 + }, + { + "epoch": 4.415465268676278, + "grad_norm": 1.7315080165863037, + "learning_rate": 3.424202176214741e-05, + "loss": 0.2682, + "num_input_tokens_seen": 4516752, + "step": 16845 + }, + { + "epoch": 4.4167758846657925, + "grad_norm": 0.8276686668395996, + "learning_rate": 3.423139374414877e-05, + "loss": 0.2936, + "num_input_tokens_seen": 4517808, + "step": 16850 + }, + { + "epoch": 4.418086500655308, + "grad_norm": 1.2451741695404053, + "learning_rate": 3.422076379403537e-05, + "loss": 0.4362, + "num_input_tokens_seen": 4518944, + "step": 16855 + }, + { + "epoch": 4.419397116644823, + "grad_norm": 0.961547315120697, + "learning_rate": 3.421013191403203e-05, + "loss": 0.2964, + "num_input_tokens_seen": 4520160, + "step": 16860 + }, + { + "epoch": 4.420707732634338, + "grad_norm": 0.9580074548721313, + "learning_rate": 3.419949810636401e-05, + "loss": 0.3035, + "num_input_tokens_seen": 4521264, + "step": 16865 + }, + { + "epoch": 4.422018348623853, + "grad_norm": 1.3352952003479004, + "learning_rate": 3.418886237325693e-05, + "loss": 0.4034, + "num_input_tokens_seen": 4522272, + "step": 16870 + }, + { + "epoch": 4.423328964613368, + "grad_norm": 1.7047370672225952, + "learning_rate": 3.417822471693683e-05, + "loss": 0.4091, + "num_input_tokens_seen": 4523344, + "step": 16875 + }, + { + "epoch": 4.424639580602883, + "grad_norm": 3.414685010910034, + "learning_rate": 3.4167585139630156e-05, + "loss": 0.3875, + "num_input_tokens_seen": 4524944, + "step": 16880 + }, + { + "epoch": 4.425950196592399, + "grad_norm": 0.7712069749832153, + "learning_rate": 3.415694364356375e-05, + "loss": 0.4026, + "num_input_tokens_seen": 4526112, + "step": 16885 + }, + { + "epoch": 4.427260812581913, + "grad_norm": 1.6919158697128296, + "learning_rate": 3.414630023096486e-05, + "loss": 0.3814, + "num_input_tokens_seen": 4527920, + "step": 16890 + }, + { + "epoch": 4.428571428571429, + "grad_norm": 1.1060981750488281, + "learning_rate": 3.413565490406114e-05, + "loss": 0.2913, + "num_input_tokens_seen": 4529280, + "step": 16895 + }, + { + "epoch": 4.4298820445609435, + "grad_norm": 1.6481386423110962, + "learning_rate": 3.412500766508063e-05, + "loss": 0.3354, + "num_input_tokens_seen": 4530736, + "step": 16900 + }, + { + "epoch": 4.431192660550459, + "grad_norm": 0.5173926949501038, + "learning_rate": 3.411435851625179e-05, + "loss": 0.4041, + "num_input_tokens_seen": 4532336, + "step": 16905 + }, + { + "epoch": 4.432503276539974, + "grad_norm": 1.5349154472351074, + "learning_rate": 3.4103707459803456e-05, + "loss": 0.3551, + "num_input_tokens_seen": 4533728, + "step": 16910 + }, + { + "epoch": 4.433813892529489, + "grad_norm": 0.9076683521270752, + "learning_rate": 3.4093054497964884e-05, + "loss": 0.368, + "num_input_tokens_seen": 4534992, + "step": 16915 + }, + { + "epoch": 4.435124508519004, + "grad_norm": 1.268669605255127, + "learning_rate": 3.40823996329657e-05, + "loss": 0.3017, + "num_input_tokens_seen": 4536512, + "step": 16920 + }, + { + "epoch": 4.436435124508519, + "grad_norm": 0.6883957386016846, + "learning_rate": 3.4071742867035985e-05, + "loss": 0.4736, + "num_input_tokens_seen": 4537984, + "step": 16925 + }, + { + "epoch": 4.437745740498034, + "grad_norm": 0.9680774211883545, + "learning_rate": 3.4061084202406154e-05, + "loss": 0.2645, + "num_input_tokens_seen": 4540032, + "step": 16930 + }, + { + "epoch": 4.43905635648755, + "grad_norm": 0.6158968210220337, + "learning_rate": 3.405042364130706e-05, + "loss": 0.2923, + "num_input_tokens_seen": 4541792, + "step": 16935 + }, + { + "epoch": 4.440366972477064, + "grad_norm": 0.8580657243728638, + "learning_rate": 3.403976118596994e-05, + "loss": 0.384, + "num_input_tokens_seen": 4542976, + "step": 16940 + }, + { + "epoch": 4.441677588466579, + "grad_norm": 1.2857410907745361, + "learning_rate": 3.4029096838626415e-05, + "loss": 0.364, + "num_input_tokens_seen": 4544368, + "step": 16945 + }, + { + "epoch": 4.442988204456094, + "grad_norm": 1.3169713020324707, + "learning_rate": 3.401843060150853e-05, + "loss": 0.3048, + "num_input_tokens_seen": 4545648, + "step": 16950 + }, + { + "epoch": 4.444298820445609, + "grad_norm": 1.1006603240966797, + "learning_rate": 3.400776247684869e-05, + "loss": 0.2863, + "num_input_tokens_seen": 4546624, + "step": 16955 + }, + { + "epoch": 4.445609436435125, + "grad_norm": 1.4565646648406982, + "learning_rate": 3.399709246687974e-05, + "loss": 0.2703, + "num_input_tokens_seen": 4547968, + "step": 16960 + }, + { + "epoch": 4.446920052424639, + "grad_norm": 0.7699347138404846, + "learning_rate": 3.398642057383488e-05, + "loss": 0.3523, + "num_input_tokens_seen": 4549872, + "step": 16965 + }, + { + "epoch": 4.448230668414155, + "grad_norm": 1.3964532613754272, + "learning_rate": 3.397574679994772e-05, + "loss": 0.2991, + "num_input_tokens_seen": 4551216, + "step": 16970 + }, + { + "epoch": 4.4495412844036695, + "grad_norm": 1.9362200498580933, + "learning_rate": 3.396507114745226e-05, + "loss": 0.3483, + "num_input_tokens_seen": 4552288, + "step": 16975 + }, + { + "epoch": 4.450851900393185, + "grad_norm": 0.8070736527442932, + "learning_rate": 3.3954393618582905e-05, + "loss": 0.3322, + "num_input_tokens_seen": 4553664, + "step": 16980 + }, + { + "epoch": 4.4521625163827, + "grad_norm": 1.471169114112854, + "learning_rate": 3.394371421557444e-05, + "loss": 0.4882, + "num_input_tokens_seen": 4555104, + "step": 16985 + }, + { + "epoch": 4.453473132372215, + "grad_norm": 1.3567899465560913, + "learning_rate": 3.3933032940662034e-05, + "loss": 0.3861, + "num_input_tokens_seen": 4556752, + "step": 16990 + }, + { + "epoch": 4.45478374836173, + "grad_norm": 1.3972630500793457, + "learning_rate": 3.392234979608128e-05, + "loss": 0.3212, + "num_input_tokens_seen": 4557728, + "step": 16995 + }, + { + "epoch": 4.456094364351245, + "grad_norm": 1.0193930864334106, + "learning_rate": 3.391166478406812e-05, + "loss": 0.251, + "num_input_tokens_seen": 4559088, + "step": 17000 + }, + { + "epoch": 4.45740498034076, + "grad_norm": 1.1194957494735718, + "learning_rate": 3.390097790685892e-05, + "loss": 0.3803, + "num_input_tokens_seen": 4560304, + "step": 17005 + }, + { + "epoch": 4.458715596330276, + "grad_norm": 0.9183666706085205, + "learning_rate": 3.3890289166690425e-05, + "loss": 0.3078, + "num_input_tokens_seen": 4561808, + "step": 17010 + }, + { + "epoch": 4.46002621231979, + "grad_norm": 0.9468809366226196, + "learning_rate": 3.387959856579977e-05, + "loss": 0.3896, + "num_input_tokens_seen": 4563984, + "step": 17015 + }, + { + "epoch": 4.461336828309306, + "grad_norm": 0.7956042289733887, + "learning_rate": 3.386890610642448e-05, + "loss": 0.5568, + "num_input_tokens_seen": 4565744, + "step": 17020 + }, + { + "epoch": 4.46264744429882, + "grad_norm": 1.2555854320526123, + "learning_rate": 3.3858211790802466e-05, + "loss": 0.311, + "num_input_tokens_seen": 4566768, + "step": 17025 + }, + { + "epoch": 4.463958060288336, + "grad_norm": 1.3201371431350708, + "learning_rate": 3.384751562117203e-05, + "loss": 0.3382, + "num_input_tokens_seen": 4567952, + "step": 17030 + }, + { + "epoch": 4.465268676277851, + "grad_norm": 0.9919450879096985, + "learning_rate": 3.3836817599771854e-05, + "loss": 0.364, + "num_input_tokens_seen": 4569120, + "step": 17035 + }, + { + "epoch": 4.466579292267365, + "grad_norm": 0.6491896510124207, + "learning_rate": 3.382611772884103e-05, + "loss": 0.4608, + "num_input_tokens_seen": 4570448, + "step": 17040 + }, + { + "epoch": 4.467889908256881, + "grad_norm": 1.5303882360458374, + "learning_rate": 3.3815416010619013e-05, + "loss": 0.3148, + "num_input_tokens_seen": 4572144, + "step": 17045 + }, + { + "epoch": 4.4692005242463955, + "grad_norm": 1.081374168395996, + "learning_rate": 3.380471244734566e-05, + "loss": 0.2438, + "num_input_tokens_seen": 4573296, + "step": 17050 + }, + { + "epoch": 4.470511140235911, + "grad_norm": 1.7200173139572144, + "learning_rate": 3.3794007041261205e-05, + "loss": 0.261, + "num_input_tokens_seen": 4574336, + "step": 17055 + }, + { + "epoch": 4.471821756225426, + "grad_norm": 1.7676063776016235, + "learning_rate": 3.378329979460627e-05, + "loss": 0.3559, + "num_input_tokens_seen": 4575840, + "step": 17060 + }, + { + "epoch": 4.473132372214941, + "grad_norm": 0.902169406414032, + "learning_rate": 3.3772590709621864e-05, + "loss": 0.256, + "num_input_tokens_seen": 4577072, + "step": 17065 + }, + { + "epoch": 4.474442988204456, + "grad_norm": 2.76840877532959, + "learning_rate": 3.376187978854938e-05, + "loss": 0.3377, + "num_input_tokens_seen": 4578384, + "step": 17070 + }, + { + "epoch": 4.475753604193971, + "grad_norm": 0.864295244216919, + "learning_rate": 3.3751167033630596e-05, + "loss": 0.2584, + "num_input_tokens_seen": 4579808, + "step": 17075 + }, + { + "epoch": 4.477064220183486, + "grad_norm": 7.900758743286133, + "learning_rate": 3.374045244710767e-05, + "loss": 0.4071, + "num_input_tokens_seen": 4580736, + "step": 17080 + }, + { + "epoch": 4.478374836173002, + "grad_norm": 1.2056097984313965, + "learning_rate": 3.372973603122316e-05, + "loss": 0.2945, + "num_input_tokens_seen": 4581792, + "step": 17085 + }, + { + "epoch": 4.479685452162516, + "grad_norm": 2.6902904510498047, + "learning_rate": 3.371901778821997e-05, + "loss": 0.3223, + "num_input_tokens_seen": 4583040, + "step": 17090 + }, + { + "epoch": 4.480996068152032, + "grad_norm": 0.6455978155136108, + "learning_rate": 3.3708297720341425e-05, + "loss": 0.3793, + "num_input_tokens_seen": 4584320, + "step": 17095 + }, + { + "epoch": 4.482306684141546, + "grad_norm": 0.9460858702659607, + "learning_rate": 3.369757582983122e-05, + "loss": 0.4335, + "num_input_tokens_seen": 4585584, + "step": 17100 + }, + { + "epoch": 4.483617300131062, + "grad_norm": 0.7704706192016602, + "learning_rate": 3.368685211893342e-05, + "loss": 0.4685, + "num_input_tokens_seen": 4587136, + "step": 17105 + }, + { + "epoch": 4.484927916120577, + "grad_norm": 0.9945593476295471, + "learning_rate": 3.367612658989247e-05, + "loss": 0.3875, + "num_input_tokens_seen": 4588624, + "step": 17110 + }, + { + "epoch": 4.486238532110092, + "grad_norm": 1.8751097917556763, + "learning_rate": 3.3665399244953236e-05, + "loss": 0.2905, + "num_input_tokens_seen": 4589632, + "step": 17115 + }, + { + "epoch": 4.487549148099607, + "grad_norm": 3.5004632472991943, + "learning_rate": 3.36546700863609e-05, + "loss": 0.3091, + "num_input_tokens_seen": 4590512, + "step": 17120 + }, + { + "epoch": 4.488859764089122, + "grad_norm": 2.038888931274414, + "learning_rate": 3.364393911636107e-05, + "loss": 0.3269, + "num_input_tokens_seen": 4591600, + "step": 17125 + }, + { + "epoch": 4.490170380078637, + "grad_norm": 0.978142261505127, + "learning_rate": 3.363320633719972e-05, + "loss": 0.3797, + "num_input_tokens_seen": 4592640, + "step": 17130 + }, + { + "epoch": 4.491480996068152, + "grad_norm": 2.4940245151519775, + "learning_rate": 3.3622471751123196e-05, + "loss": 0.5637, + "num_input_tokens_seen": 4594224, + "step": 17135 + }, + { + "epoch": 4.492791612057667, + "grad_norm": 0.8812064528465271, + "learning_rate": 3.361173536037824e-05, + "loss": 0.2937, + "num_input_tokens_seen": 4595808, + "step": 17140 + }, + { + "epoch": 4.494102228047182, + "grad_norm": 2.1827640533447266, + "learning_rate": 3.360099716721195e-05, + "loss": 0.3161, + "num_input_tokens_seen": 4597264, + "step": 17145 + }, + { + "epoch": 4.495412844036697, + "grad_norm": 1.1164640188217163, + "learning_rate": 3.3590257173871807e-05, + "loss": 0.3574, + "num_input_tokens_seen": 4598544, + "step": 17150 + }, + { + "epoch": 4.496723460026212, + "grad_norm": 1.292435646057129, + "learning_rate": 3.357951538260569e-05, + "loss": 0.433, + "num_input_tokens_seen": 4599728, + "step": 17155 + }, + { + "epoch": 4.498034076015728, + "grad_norm": 0.8596558570861816, + "learning_rate": 3.356877179566183e-05, + "loss": 0.3047, + "num_input_tokens_seen": 4601216, + "step": 17160 + }, + { + "epoch": 4.499344692005242, + "grad_norm": 1.2215769290924072, + "learning_rate": 3.355802641528883e-05, + "loss": 0.4204, + "num_input_tokens_seen": 4602576, + "step": 17165 + }, + { + "epoch": 4.500655307994758, + "grad_norm": 0.9570655226707458, + "learning_rate": 3.3547279243735695e-05, + "loss": 0.3171, + "num_input_tokens_seen": 4603984, + "step": 17170 + }, + { + "epoch": 4.501179554390563, + "eval_loss": 0.5279697775840759, + "eval_runtime": 16.8241, + "eval_samples_per_second": 50.404, + "eval_steps_per_second": 25.202, + "num_input_tokens_seen": 4604416, + "step": 17172 + }, + { + "epoch": 4.501965923984272, + "grad_norm": 0.4432520270347595, + "learning_rate": 3.3536530283251784e-05, + "loss": 0.3837, + "num_input_tokens_seen": 4605232, + "step": 17175 + }, + { + "epoch": 4.503276539973788, + "grad_norm": 0.48370397090911865, + "learning_rate": 3.3525779536086826e-05, + "loss": 0.2866, + "num_input_tokens_seen": 4606368, + "step": 17180 + }, + { + "epoch": 4.504587155963303, + "grad_norm": 1.056260108947754, + "learning_rate": 3.351502700449095e-05, + "loss": 0.4816, + "num_input_tokens_seen": 4607600, + "step": 17185 + }, + { + "epoch": 4.505897771952818, + "grad_norm": 2.0200674533843994, + "learning_rate": 3.350427269071463e-05, + "loss": 0.3181, + "num_input_tokens_seen": 4608800, + "step": 17190 + }, + { + "epoch": 4.507208387942333, + "grad_norm": 1.6030006408691406, + "learning_rate": 3.3493516597008734e-05, + "loss": 0.3577, + "num_input_tokens_seen": 4609696, + "step": 17195 + }, + { + "epoch": 4.508519003931848, + "grad_norm": 0.7797363996505737, + "learning_rate": 3.348275872562449e-05, + "loss": 0.2513, + "num_input_tokens_seen": 4610944, + "step": 17200 + }, + { + "epoch": 4.509829619921363, + "grad_norm": 0.9738597273826599, + "learning_rate": 3.34719990788135e-05, + "loss": 0.3427, + "num_input_tokens_seen": 4612368, + "step": 17205 + }, + { + "epoch": 4.511140235910878, + "grad_norm": 1.6520428657531738, + "learning_rate": 3.346123765882774e-05, + "loss": 0.2995, + "num_input_tokens_seen": 4613472, + "step": 17210 + }, + { + "epoch": 4.512450851900393, + "grad_norm": 1.000096321105957, + "learning_rate": 3.345047446791956e-05, + "loss": 0.4784, + "num_input_tokens_seen": 4614688, + "step": 17215 + }, + { + "epoch": 4.513761467889909, + "grad_norm": 0.8260378837585449, + "learning_rate": 3.3439709508341676e-05, + "loss": 0.3078, + "num_input_tokens_seen": 4616160, + "step": 17220 + }, + { + "epoch": 4.515072083879423, + "grad_norm": 0.511508584022522, + "learning_rate": 3.342894278234717e-05, + "loss": 0.2536, + "num_input_tokens_seen": 4617264, + "step": 17225 + }, + { + "epoch": 4.516382699868938, + "grad_norm": 0.7919830083847046, + "learning_rate": 3.341817429218951e-05, + "loss": 0.3764, + "num_input_tokens_seen": 4619456, + "step": 17230 + }, + { + "epoch": 4.517693315858454, + "grad_norm": 0.7993356585502625, + "learning_rate": 3.340740404012251e-05, + "loss": 0.2881, + "num_input_tokens_seen": 4620960, + "step": 17235 + }, + { + "epoch": 4.519003931847968, + "grad_norm": 0.8063624501228333, + "learning_rate": 3.339663202840038e-05, + "loss": 0.2579, + "num_input_tokens_seen": 4622176, + "step": 17240 + }, + { + "epoch": 4.520314547837484, + "grad_norm": 1.8179179430007935, + "learning_rate": 3.3385858259277666e-05, + "loss": 0.37, + "num_input_tokens_seen": 4623808, + "step": 17245 + }, + { + "epoch": 4.521625163826998, + "grad_norm": 2.0231552124023438, + "learning_rate": 3.337508273500931e-05, + "loss": 0.404, + "num_input_tokens_seen": 4626640, + "step": 17250 + }, + { + "epoch": 4.522935779816514, + "grad_norm": 1.1218013763427734, + "learning_rate": 3.336430545785061e-05, + "loss": 0.2523, + "num_input_tokens_seen": 4627824, + "step": 17255 + }, + { + "epoch": 4.524246395806029, + "grad_norm": 0.65257328748703, + "learning_rate": 3.335352643005723e-05, + "loss": 0.2844, + "num_input_tokens_seen": 4629232, + "step": 17260 + }, + { + "epoch": 4.525557011795544, + "grad_norm": 1.2763421535491943, + "learning_rate": 3.33427456538852e-05, + "loss": 0.4962, + "num_input_tokens_seen": 4630416, + "step": 17265 + }, + { + "epoch": 4.526867627785059, + "grad_norm": 0.8579710721969604, + "learning_rate": 3.3331963131590916e-05, + "loss": 0.4714, + "num_input_tokens_seen": 4631984, + "step": 17270 + }, + { + "epoch": 4.528178243774574, + "grad_norm": 0.8161243200302124, + "learning_rate": 3.3321178865431155e-05, + "loss": 0.2323, + "num_input_tokens_seen": 4633104, + "step": 17275 + }, + { + "epoch": 4.529488859764089, + "grad_norm": 1.0353360176086426, + "learning_rate": 3.331039285766303e-05, + "loss": 0.4396, + "num_input_tokens_seen": 4634656, + "step": 17280 + }, + { + "epoch": 4.5307994757536045, + "grad_norm": 1.4839775562286377, + "learning_rate": 3.329960511054404e-05, + "loss": 0.4718, + "num_input_tokens_seen": 4635616, + "step": 17285 + }, + { + "epoch": 4.532110091743119, + "grad_norm": 1.3539988994598389, + "learning_rate": 3.3288815626332035e-05, + "loss": 0.2231, + "num_input_tokens_seen": 4636720, + "step": 17290 + }, + { + "epoch": 4.533420707732635, + "grad_norm": 2.9153738021850586, + "learning_rate": 3.327802440728526e-05, + "loss": 0.389, + "num_input_tokens_seen": 4637904, + "step": 17295 + }, + { + "epoch": 4.534731323722149, + "grad_norm": 2.379587411880493, + "learning_rate": 3.3267231455662265e-05, + "loss": 0.3614, + "num_input_tokens_seen": 4639264, + "step": 17300 + }, + { + "epoch": 4.536041939711664, + "grad_norm": 1.3473591804504395, + "learning_rate": 3.325643677372201e-05, + "loss": 0.2648, + "num_input_tokens_seen": 4640176, + "step": 17305 + }, + { + "epoch": 4.53735255570118, + "grad_norm": 1.3348702192306519, + "learning_rate": 3.324564036372381e-05, + "loss": 0.4249, + "num_input_tokens_seen": 4641440, + "step": 17310 + }, + { + "epoch": 4.538663171690695, + "grad_norm": 1.2591218948364258, + "learning_rate": 3.323484222792732e-05, + "loss": 0.3616, + "num_input_tokens_seen": 4642496, + "step": 17315 + }, + { + "epoch": 4.53997378768021, + "grad_norm": 0.9491422176361084, + "learning_rate": 3.322404236859259e-05, + "loss": 0.4466, + "num_input_tokens_seen": 4643744, + "step": 17320 + }, + { + "epoch": 4.541284403669724, + "grad_norm": 3.3169338703155518, + "learning_rate": 3.321324078798e-05, + "loss": 0.2892, + "num_input_tokens_seen": 4645024, + "step": 17325 + }, + { + "epoch": 4.54259501965924, + "grad_norm": 0.9081875085830688, + "learning_rate": 3.32024374883503e-05, + "loss": 0.2813, + "num_input_tokens_seen": 4645984, + "step": 17330 + }, + { + "epoch": 4.543905635648755, + "grad_norm": 1.8820483684539795, + "learning_rate": 3.319163247196462e-05, + "loss": 0.4431, + "num_input_tokens_seen": 4647072, + "step": 17335 + }, + { + "epoch": 4.54521625163827, + "grad_norm": 0.66603684425354, + "learning_rate": 3.318082574108441e-05, + "loss": 0.3084, + "num_input_tokens_seen": 4648304, + "step": 17340 + }, + { + "epoch": 4.546526867627785, + "grad_norm": 0.9968544840812683, + "learning_rate": 3.31700172979715e-05, + "loss": 0.305, + "num_input_tokens_seen": 4649728, + "step": 17345 + }, + { + "epoch": 4.5478374836173, + "grad_norm": 0.8024002313613892, + "learning_rate": 3.315920714488808e-05, + "loss": 0.273, + "num_input_tokens_seen": 4650816, + "step": 17350 + }, + { + "epoch": 4.549148099606815, + "grad_norm": 1.6931535005569458, + "learning_rate": 3.314839528409672e-05, + "loss": 0.4184, + "num_input_tokens_seen": 4651888, + "step": 17355 + }, + { + "epoch": 4.5504587155963305, + "grad_norm": 1.4065825939178467, + "learning_rate": 3.313758171786031e-05, + "loss": 0.4103, + "num_input_tokens_seen": 4653008, + "step": 17360 + }, + { + "epoch": 4.551769331585845, + "grad_norm": 1.1847862005233765, + "learning_rate": 3.3126766448442095e-05, + "loss": 0.4605, + "num_input_tokens_seen": 4654560, + "step": 17365 + }, + { + "epoch": 4.553079947575361, + "grad_norm": 0.8436602354049683, + "learning_rate": 3.311594947810571e-05, + "loss": 0.3433, + "num_input_tokens_seen": 4655552, + "step": 17370 + }, + { + "epoch": 4.554390563564875, + "grad_norm": 1.8319445848464966, + "learning_rate": 3.310513080911512e-05, + "loss": 0.5154, + "num_input_tokens_seen": 4656800, + "step": 17375 + }, + { + "epoch": 4.555701179554391, + "grad_norm": 0.8904233574867249, + "learning_rate": 3.309431044373465e-05, + "loss": 0.3977, + "num_input_tokens_seen": 4658208, + "step": 17380 + }, + { + "epoch": 4.557011795543906, + "grad_norm": 18.091218948364258, + "learning_rate": 3.3083488384228995e-05, + "loss": 0.2255, + "num_input_tokens_seen": 4659248, + "step": 17385 + }, + { + "epoch": 4.558322411533421, + "grad_norm": 0.9263575673103333, + "learning_rate": 3.3072664632863195e-05, + "loss": 0.3152, + "num_input_tokens_seen": 4660480, + "step": 17390 + }, + { + "epoch": 4.559633027522936, + "grad_norm": 1.5175762176513672, + "learning_rate": 3.3061839191902635e-05, + "loss": 0.3314, + "num_input_tokens_seen": 4661408, + "step": 17395 + }, + { + "epoch": 4.56094364351245, + "grad_norm": 0.7315385341644287, + "learning_rate": 3.3051012063613054e-05, + "loss": 0.2387, + "num_input_tokens_seen": 4662464, + "step": 17400 + }, + { + "epoch": 4.562254259501966, + "grad_norm": 1.861042857170105, + "learning_rate": 3.3040183250260556e-05, + "loss": 0.4977, + "num_input_tokens_seen": 4663744, + "step": 17405 + }, + { + "epoch": 4.5635648754914815, + "grad_norm": 1.1015620231628418, + "learning_rate": 3.30293527541116e-05, + "loss": 0.3673, + "num_input_tokens_seen": 4665008, + "step": 17410 + }, + { + "epoch": 4.564875491480996, + "grad_norm": 0.336713969707489, + "learning_rate": 3.3018520577432996e-05, + "loss": 0.3312, + "num_input_tokens_seen": 4666672, + "step": 17415 + }, + { + "epoch": 4.566186107470511, + "grad_norm": 1.021867036819458, + "learning_rate": 3.3007686722491885e-05, + "loss": 0.3048, + "num_input_tokens_seen": 4667552, + "step": 17420 + }, + { + "epoch": 4.567496723460026, + "grad_norm": 0.6166992783546448, + "learning_rate": 3.299685119155577e-05, + "loss": 0.418, + "num_input_tokens_seen": 4668592, + "step": 17425 + }, + { + "epoch": 4.568807339449541, + "grad_norm": 1.4782077074050903, + "learning_rate": 3.298601398689253e-05, + "loss": 0.2386, + "num_input_tokens_seen": 4669984, + "step": 17430 + }, + { + "epoch": 4.5701179554390565, + "grad_norm": 1.722707748413086, + "learning_rate": 3.297517511077035e-05, + "loss": 0.5244, + "num_input_tokens_seen": 4671168, + "step": 17435 + }, + { + "epoch": 4.571428571428571, + "grad_norm": 1.460835337638855, + "learning_rate": 3.296433456545781e-05, + "loss": 0.3672, + "num_input_tokens_seen": 4672768, + "step": 17440 + }, + { + "epoch": 4.572739187418087, + "grad_norm": 0.837135910987854, + "learning_rate": 3.2953492353223805e-05, + "loss": 0.3425, + "num_input_tokens_seen": 4674544, + "step": 17445 + }, + { + "epoch": 4.574049803407601, + "grad_norm": 1.9367479085922241, + "learning_rate": 3.294264847633759e-05, + "loss": 0.4289, + "num_input_tokens_seen": 4675584, + "step": 17450 + }, + { + "epoch": 4.575360419397117, + "grad_norm": 1.7259907722473145, + "learning_rate": 3.2931802937068776e-05, + "loss": 0.6242, + "num_input_tokens_seen": 4676784, + "step": 17455 + }, + { + "epoch": 4.576671035386632, + "grad_norm": 0.9007288813591003, + "learning_rate": 3.292095573768731e-05, + "loss": 0.3184, + "num_input_tokens_seen": 4677856, + "step": 17460 + }, + { + "epoch": 4.577981651376147, + "grad_norm": 0.8866837024688721, + "learning_rate": 3.2910106880463486e-05, + "loss": 0.315, + "num_input_tokens_seen": 4679040, + "step": 17465 + }, + { + "epoch": 4.579292267365662, + "grad_norm": 2.9022462368011475, + "learning_rate": 3.289925636766797e-05, + "loss": 0.3083, + "num_input_tokens_seen": 4680400, + "step": 17470 + }, + { + "epoch": 4.580602883355177, + "grad_norm": 2.580007553100586, + "learning_rate": 3.2888404201571743e-05, + "loss": 0.3716, + "num_input_tokens_seen": 4681888, + "step": 17475 + }, + { + "epoch": 4.581913499344692, + "grad_norm": 1.1763910055160522, + "learning_rate": 3.2877550384446146e-05, + "loss": 0.3805, + "num_input_tokens_seen": 4682848, + "step": 17480 + }, + { + "epoch": 4.5832241153342075, + "grad_norm": 1.625781536102295, + "learning_rate": 3.2866694918562866e-05, + "loss": 0.3764, + "num_input_tokens_seen": 4684096, + "step": 17485 + }, + { + "epoch": 4.584534731323722, + "grad_norm": 0.5062465667724609, + "learning_rate": 3.285583780619394e-05, + "loss": 0.2697, + "num_input_tokens_seen": 4685344, + "step": 17490 + }, + { + "epoch": 4.585845347313237, + "grad_norm": 0.6561262011528015, + "learning_rate": 3.284497904961172e-05, + "loss": 0.3557, + "num_input_tokens_seen": 4686768, + "step": 17495 + }, + { + "epoch": 4.587155963302752, + "grad_norm": 1.8110065460205078, + "learning_rate": 3.283411865108894e-05, + "loss": 0.4009, + "num_input_tokens_seen": 4687776, + "step": 17500 + }, + { + "epoch": 4.588466579292267, + "grad_norm": 1.43059504032135, + "learning_rate": 3.2823256612898676e-05, + "loss": 0.3891, + "num_input_tokens_seen": 4688752, + "step": 17505 + }, + { + "epoch": 4.5897771952817825, + "grad_norm": 1.6707801818847656, + "learning_rate": 3.2812392937314316e-05, + "loss": 0.3017, + "num_input_tokens_seen": 4690128, + "step": 17510 + }, + { + "epoch": 4.591087811271297, + "grad_norm": 2.7570295333862305, + "learning_rate": 3.2801527626609606e-05, + "loss": 0.3322, + "num_input_tokens_seen": 4691312, + "step": 17515 + }, + { + "epoch": 4.592398427260813, + "grad_norm": 1.3228874206542969, + "learning_rate": 3.2790660683058646e-05, + "loss": 0.3666, + "num_input_tokens_seen": 4693120, + "step": 17520 + }, + { + "epoch": 4.593709043250327, + "grad_norm": 0.6064137816429138, + "learning_rate": 3.277979210893587e-05, + "loss": 0.3553, + "num_input_tokens_seen": 4694640, + "step": 17525 + }, + { + "epoch": 4.595019659239843, + "grad_norm": 0.43361860513687134, + "learning_rate": 3.276892190651605e-05, + "loss": 0.2765, + "num_input_tokens_seen": 4695888, + "step": 17530 + }, + { + "epoch": 4.5963302752293576, + "grad_norm": 1.843876838684082, + "learning_rate": 3.275805007807429e-05, + "loss": 0.3136, + "num_input_tokens_seen": 4696928, + "step": 17535 + }, + { + "epoch": 4.597640891218873, + "grad_norm": 0.9966059923171997, + "learning_rate": 3.274717662588606e-05, + "loss": 0.4569, + "num_input_tokens_seen": 4698608, + "step": 17540 + }, + { + "epoch": 4.598951507208388, + "grad_norm": 0.7497085928916931, + "learning_rate": 3.273630155222715e-05, + "loss": 0.2121, + "num_input_tokens_seen": 4699856, + "step": 17545 + }, + { + "epoch": 4.600262123197903, + "grad_norm": 1.582413911819458, + "learning_rate": 3.272542485937369e-05, + "loss": 0.326, + "num_input_tokens_seen": 4701488, + "step": 17550 + }, + { + "epoch": 4.601572739187418, + "grad_norm": 1.3884965181350708, + "learning_rate": 3.271454654960215e-05, + "loss": 0.4373, + "num_input_tokens_seen": 4702960, + "step": 17555 + }, + { + "epoch": 4.6028833551769335, + "grad_norm": 2.1121835708618164, + "learning_rate": 3.270366662518936e-05, + "loss": 0.276, + "num_input_tokens_seen": 4704208, + "step": 17560 + }, + { + "epoch": 4.604193971166448, + "grad_norm": 1.6815704107284546, + "learning_rate": 3.2692785088412444e-05, + "loss": 0.1696, + "num_input_tokens_seen": 4705440, + "step": 17565 + }, + { + "epoch": 4.605504587155964, + "grad_norm": 2.9938669204711914, + "learning_rate": 3.268190194154891e-05, + "loss": 0.3825, + "num_input_tokens_seen": 4706944, + "step": 17570 + }, + { + "epoch": 4.606815203145478, + "grad_norm": 3.044672727584839, + "learning_rate": 3.267101718687657e-05, + "loss": 0.3459, + "num_input_tokens_seen": 4708304, + "step": 17575 + }, + { + "epoch": 4.608125819134994, + "grad_norm": 0.6499961614608765, + "learning_rate": 3.266013082667357e-05, + "loss": 0.1923, + "num_input_tokens_seen": 4709424, + "step": 17580 + }, + { + "epoch": 4.6094364351245085, + "grad_norm": 2.7903130054473877, + "learning_rate": 3.2649242863218446e-05, + "loss": 0.3787, + "num_input_tokens_seen": 4710832, + "step": 17585 + }, + { + "epoch": 4.610747051114023, + "grad_norm": 1.091233730316162, + "learning_rate": 3.2638353298790005e-05, + "loss": 0.332, + "num_input_tokens_seen": 4712144, + "step": 17590 + }, + { + "epoch": 4.612057667103539, + "grad_norm": 1.0650490522384644, + "learning_rate": 3.2627462135667405e-05, + "loss": 0.4826, + "num_input_tokens_seen": 4713648, + "step": 17595 + }, + { + "epoch": 4.613368283093053, + "grad_norm": 1.1756516695022583, + "learning_rate": 3.2616569376130175e-05, + "loss": 0.2266, + "num_input_tokens_seen": 4715168, + "step": 17600 + }, + { + "epoch": 4.614678899082569, + "grad_norm": 1.6492366790771484, + "learning_rate": 3.260567502245813e-05, + "loss": 0.5748, + "num_input_tokens_seen": 4716272, + "step": 17605 + }, + { + "epoch": 4.6159895150720835, + "grad_norm": 1.0576963424682617, + "learning_rate": 3.259477907693144e-05, + "loss": 0.3167, + "num_input_tokens_seen": 4717504, + "step": 17610 + }, + { + "epoch": 4.617300131061599, + "grad_norm": 1.2394706010818481, + "learning_rate": 3.2583881541830604e-05, + "loss": 0.3883, + "num_input_tokens_seen": 4718768, + "step": 17615 + }, + { + "epoch": 4.618610747051114, + "grad_norm": 2.204073905944824, + "learning_rate": 3.257298241943648e-05, + "loss": 0.3418, + "num_input_tokens_seen": 4720160, + "step": 17620 + }, + { + "epoch": 4.619921363040629, + "grad_norm": 2.110076904296875, + "learning_rate": 3.256208171203021e-05, + "loss": 0.359, + "num_input_tokens_seen": 4721168, + "step": 17625 + }, + { + "epoch": 4.621231979030144, + "grad_norm": 1.7945225238800049, + "learning_rate": 3.255117942189331e-05, + "loss": 0.3088, + "num_input_tokens_seen": 4722304, + "step": 17630 + }, + { + "epoch": 4.6225425950196595, + "grad_norm": 9.34453010559082, + "learning_rate": 3.2540275551307605e-05, + "loss": 0.2622, + "num_input_tokens_seen": 4723792, + "step": 17635 + }, + { + "epoch": 4.623853211009174, + "grad_norm": 2.40902042388916, + "learning_rate": 3.252937010255525e-05, + "loss": 0.3361, + "num_input_tokens_seen": 4725312, + "step": 17640 + }, + { + "epoch": 4.62516382699869, + "grad_norm": 1.8108171224594116, + "learning_rate": 3.2518463077918746e-05, + "loss": 0.2513, + "num_input_tokens_seen": 4726400, + "step": 17645 + }, + { + "epoch": 4.626474442988204, + "grad_norm": 0.878288984298706, + "learning_rate": 3.250755447968091e-05, + "loss": 0.3911, + "num_input_tokens_seen": 4728416, + "step": 17650 + }, + { + "epoch": 4.62778505897772, + "grad_norm": 2.2596187591552734, + "learning_rate": 3.24966443101249e-05, + "loss": 0.4441, + "num_input_tokens_seen": 4729472, + "step": 17655 + }, + { + "epoch": 4.6290956749672345, + "grad_norm": 0.8037793040275574, + "learning_rate": 3.248573257153418e-05, + "loss": 0.4202, + "num_input_tokens_seen": 4730912, + "step": 17660 + }, + { + "epoch": 4.63040629095675, + "grad_norm": 1.7311135530471802, + "learning_rate": 3.2474819266192574e-05, + "loss": 0.2136, + "num_input_tokens_seen": 4732064, + "step": 17665 + }, + { + "epoch": 4.631716906946265, + "grad_norm": 0.7242838740348816, + "learning_rate": 3.24639043963842e-05, + "loss": 0.3932, + "num_input_tokens_seen": 4733536, + "step": 17670 + }, + { + "epoch": 4.63302752293578, + "grad_norm": 0.8664498329162598, + "learning_rate": 3.245298796439354e-05, + "loss": 0.1977, + "num_input_tokens_seen": 4734992, + "step": 17675 + }, + { + "epoch": 4.634338138925295, + "grad_norm": 1.0368537902832031, + "learning_rate": 3.244206997250538e-05, + "loss": 0.3628, + "num_input_tokens_seen": 4736480, + "step": 17680 + }, + { + "epoch": 4.6356487549148095, + "grad_norm": 1.984461784362793, + "learning_rate": 3.243115042300483e-05, + "loss": 0.3232, + "num_input_tokens_seen": 4737936, + "step": 17685 + }, + { + "epoch": 4.636959370904325, + "grad_norm": 1.210694670677185, + "learning_rate": 3.242022931817734e-05, + "loss": 0.3259, + "num_input_tokens_seen": 4739088, + "step": 17690 + }, + { + "epoch": 4.63826998689384, + "grad_norm": 1.4179786443710327, + "learning_rate": 3.240930666030867e-05, + "loss": 0.3469, + "num_input_tokens_seen": 4740352, + "step": 17695 + }, + { + "epoch": 4.639580602883355, + "grad_norm": 1.5681769847869873, + "learning_rate": 3.239838245168492e-05, + "loss": 0.3541, + "num_input_tokens_seen": 4741520, + "step": 17700 + }, + { + "epoch": 4.64089121887287, + "grad_norm": 1.8151805400848389, + "learning_rate": 3.238745669459251e-05, + "loss": 0.3272, + "num_input_tokens_seen": 4742672, + "step": 17705 + }, + { + "epoch": 4.6422018348623855, + "grad_norm": 1.0693745613098145, + "learning_rate": 3.237652939131818e-05, + "loss": 0.4202, + "num_input_tokens_seen": 4743920, + "step": 17710 + }, + { + "epoch": 4.6435124508519, + "grad_norm": 0.7988050580024719, + "learning_rate": 3.236560054414899e-05, + "loss": 0.3645, + "num_input_tokens_seen": 4745472, + "step": 17715 + }, + { + "epoch": 4.644823066841416, + "grad_norm": 1.8314697742462158, + "learning_rate": 3.2354670155372326e-05, + "loss": 0.5206, + "num_input_tokens_seen": 4746528, + "step": 17720 + }, + { + "epoch": 4.64613368283093, + "grad_norm": 1.2069045305252075, + "learning_rate": 3.234373822727591e-05, + "loss": 0.5343, + "num_input_tokens_seen": 4747744, + "step": 17725 + }, + { + "epoch": 4.647444298820446, + "grad_norm": 1.811295986175537, + "learning_rate": 3.233280476214778e-05, + "loss": 0.2836, + "num_input_tokens_seen": 4748992, + "step": 17730 + }, + { + "epoch": 4.6487549148099605, + "grad_norm": 1.2789314985275269, + "learning_rate": 3.232186976227628e-05, + "loss": 0.4292, + "num_input_tokens_seen": 4750240, + "step": 17735 + }, + { + "epoch": 4.650065530799476, + "grad_norm": 1.099923849105835, + "learning_rate": 3.231093322995008e-05, + "loss": 0.3531, + "num_input_tokens_seen": 4751376, + "step": 17740 + }, + { + "epoch": 4.651376146788991, + "grad_norm": 0.9576451182365417, + "learning_rate": 3.2299995167458196e-05, + "loss": 0.2668, + "num_input_tokens_seen": 4752720, + "step": 17745 + }, + { + "epoch": 4.652686762778506, + "grad_norm": 1.064269781112671, + "learning_rate": 3.2289055577089936e-05, + "loss": 0.4106, + "num_input_tokens_seen": 4754144, + "step": 17750 + }, + { + "epoch": 4.653997378768021, + "grad_norm": 1.1905009746551514, + "learning_rate": 3.227811446113494e-05, + "loss": 0.4221, + "num_input_tokens_seen": 4755568, + "step": 17755 + }, + { + "epoch": 4.655307994757536, + "grad_norm": 0.565153181552887, + "learning_rate": 3.226717182188316e-05, + "loss": 0.4273, + "num_input_tokens_seen": 4756816, + "step": 17760 + }, + { + "epoch": 4.656618610747051, + "grad_norm": 1.5141183137893677, + "learning_rate": 3.2256227661624874e-05, + "loss": 0.5992, + "num_input_tokens_seen": 4758080, + "step": 17765 + }, + { + "epoch": 4.657929226736567, + "grad_norm": 1.574083685874939, + "learning_rate": 3.2245281982650683e-05, + "loss": 0.5591, + "num_input_tokens_seen": 4759248, + "step": 17770 + }, + { + "epoch": 4.659239842726081, + "grad_norm": 1.0873351097106934, + "learning_rate": 3.2234334787251487e-05, + "loss": 0.4024, + "num_input_tokens_seen": 4760720, + "step": 17775 + }, + { + "epoch": 4.660550458715596, + "grad_norm": 1.113553524017334, + "learning_rate": 3.222338607771851e-05, + "loss": 0.3253, + "num_input_tokens_seen": 4762256, + "step": 17780 + }, + { + "epoch": 4.6618610747051115, + "grad_norm": 1.4343210458755493, + "learning_rate": 3.221243585634333e-05, + "loss": 0.4095, + "num_input_tokens_seen": 4763632, + "step": 17785 + }, + { + "epoch": 4.663171690694626, + "grad_norm": 0.8032976984977722, + "learning_rate": 3.2201484125417766e-05, + "loss": 0.2255, + "num_input_tokens_seen": 4765456, + "step": 17790 + }, + { + "epoch": 4.664482306684142, + "grad_norm": 2.9637856483459473, + "learning_rate": 3.219053088723403e-05, + "loss": 0.3291, + "num_input_tokens_seen": 4766400, + "step": 17795 + }, + { + "epoch": 4.665792922673656, + "grad_norm": 1.0207997560501099, + "learning_rate": 3.21795761440846e-05, + "loss": 0.4237, + "num_input_tokens_seen": 4767552, + "step": 17800 + }, + { + "epoch": 4.667103538663172, + "grad_norm": 2.0791587829589844, + "learning_rate": 3.2168619898262294e-05, + "loss": 0.3564, + "num_input_tokens_seen": 4768656, + "step": 17805 + }, + { + "epoch": 4.6684141546526865, + "grad_norm": 1.3373984098434448, + "learning_rate": 3.215766215206022e-05, + "loss": 0.5361, + "num_input_tokens_seen": 4770288, + "step": 17810 + }, + { + "epoch": 4.669724770642202, + "grad_norm": 1.7787818908691406, + "learning_rate": 3.2146702907771824e-05, + "loss": 0.4249, + "num_input_tokens_seen": 4771744, + "step": 17815 + }, + { + "epoch": 4.671035386631717, + "grad_norm": 0.9821608066558838, + "learning_rate": 3.2135742167690866e-05, + "loss": 0.3876, + "num_input_tokens_seen": 4773232, + "step": 17820 + }, + { + "epoch": 4.672346002621232, + "grad_norm": 0.8207338452339172, + "learning_rate": 3.212477993411141e-05, + "loss": 0.635, + "num_input_tokens_seen": 4774832, + "step": 17825 + }, + { + "epoch": 4.673656618610747, + "grad_norm": 1.6061018705368042, + "learning_rate": 3.211381620932781e-05, + "loss": 0.4328, + "num_input_tokens_seen": 4776352, + "step": 17830 + }, + { + "epoch": 4.674967234600262, + "grad_norm": 0.5976434946060181, + "learning_rate": 3.2102850995634765e-05, + "loss": 0.379, + "num_input_tokens_seen": 4777968, + "step": 17835 + }, + { + "epoch": 4.676277850589777, + "grad_norm": 0.7899029850959778, + "learning_rate": 3.209188429532729e-05, + "loss": 0.3511, + "num_input_tokens_seen": 4779248, + "step": 17840 + }, + { + "epoch": 4.677588466579293, + "grad_norm": 0.8821409940719604, + "learning_rate": 3.2080916110700684e-05, + "loss": 0.2264, + "num_input_tokens_seen": 4780368, + "step": 17845 + }, + { + "epoch": 4.678899082568807, + "grad_norm": 0.8742518424987793, + "learning_rate": 3.206994644405057e-05, + "loss": 0.4027, + "num_input_tokens_seen": 4782304, + "step": 17850 + }, + { + "epoch": 4.680209698558322, + "grad_norm": 4.053839206695557, + "learning_rate": 3.2058975297672874e-05, + "loss": 0.3022, + "num_input_tokens_seen": 4783232, + "step": 17855 + }, + { + "epoch": 4.6815203145478375, + "grad_norm": 0.6377301812171936, + "learning_rate": 3.204800267386385e-05, + "loss": 0.304, + "num_input_tokens_seen": 4784448, + "step": 17860 + }, + { + "epoch": 4.682830930537353, + "grad_norm": 0.866705596446991, + "learning_rate": 3.203702857492005e-05, + "loss": 0.2507, + "num_input_tokens_seen": 4785376, + "step": 17865 + }, + { + "epoch": 4.684141546526868, + "grad_norm": 1.3488157987594604, + "learning_rate": 3.202605300313833e-05, + "loss": 0.4365, + "num_input_tokens_seen": 4786992, + "step": 17870 + }, + { + "epoch": 4.685452162516382, + "grad_norm": 0.8765212893486023, + "learning_rate": 3.201507596081585e-05, + "loss": 0.2885, + "num_input_tokens_seen": 4787920, + "step": 17875 + }, + { + "epoch": 4.686762778505898, + "grad_norm": 1.6845372915267944, + "learning_rate": 3.200409745025009e-05, + "loss": 0.287, + "num_input_tokens_seen": 4789392, + "step": 17880 + }, + { + "epoch": 4.6880733944954125, + "grad_norm": 0.7798876762390137, + "learning_rate": 3.199311747373885e-05, + "loss": 0.4524, + "num_input_tokens_seen": 4791888, + "step": 17885 + }, + { + "epoch": 4.689384010484928, + "grad_norm": 1.0553714036941528, + "learning_rate": 3.198213603358019e-05, + "loss": 0.2922, + "num_input_tokens_seen": 4793008, + "step": 17890 + }, + { + "epoch": 4.690694626474443, + "grad_norm": 1.1405469179153442, + "learning_rate": 3.197115313207252e-05, + "loss": 0.2953, + "num_input_tokens_seen": 4794272, + "step": 17895 + }, + { + "epoch": 4.692005242463958, + "grad_norm": 1.0685025453567505, + "learning_rate": 3.1960168771514565e-05, + "loss": 0.4123, + "num_input_tokens_seen": 4795552, + "step": 17900 + }, + { + "epoch": 4.693315858453473, + "grad_norm": 1.3015272617340088, + "learning_rate": 3.1949182954205295e-05, + "loss": 0.2065, + "num_input_tokens_seen": 4796832, + "step": 17905 + }, + { + "epoch": 4.694626474442988, + "grad_norm": 0.9997621178627014, + "learning_rate": 3.193819568244405e-05, + "loss": 0.4119, + "num_input_tokens_seen": 4798432, + "step": 17910 + }, + { + "epoch": 4.695937090432503, + "grad_norm": 1.0663864612579346, + "learning_rate": 3.192720695853043e-05, + "loss": 0.3347, + "num_input_tokens_seen": 4799792, + "step": 17915 + }, + { + "epoch": 4.697247706422019, + "grad_norm": 1.9948105812072754, + "learning_rate": 3.1916216784764365e-05, + "loss": 0.3941, + "num_input_tokens_seen": 4800864, + "step": 17920 + }, + { + "epoch": 4.698558322411533, + "grad_norm": 1.221804141998291, + "learning_rate": 3.1905225163446075e-05, + "loss": 0.3232, + "num_input_tokens_seen": 4802176, + "step": 17925 + }, + { + "epoch": 4.699868938401049, + "grad_norm": 1.480920672416687, + "learning_rate": 3.189423209687609e-05, + "loss": 0.4085, + "num_input_tokens_seen": 4803568, + "step": 17930 + }, + { + "epoch": 4.7011795543905635, + "grad_norm": 1.1945922374725342, + "learning_rate": 3.188323758735524e-05, + "loss": 0.3275, + "num_input_tokens_seen": 4804784, + "step": 17935 + }, + { + "epoch": 4.702490170380079, + "grad_norm": 1.3174827098846436, + "learning_rate": 3.187224163718466e-05, + "loss": 0.4444, + "num_input_tokens_seen": 4805920, + "step": 17940 + }, + { + "epoch": 4.703800786369594, + "grad_norm": 0.9522801637649536, + "learning_rate": 3.186124424866578e-05, + "loss": 0.3112, + "num_input_tokens_seen": 4807472, + "step": 17945 + }, + { + "epoch": 4.705111402359108, + "grad_norm": 1.506898045539856, + "learning_rate": 3.185024542410033e-05, + "loss": 0.3725, + "num_input_tokens_seen": 4808912, + "step": 17950 + }, + { + "epoch": 4.706422018348624, + "grad_norm": 0.9967443346977234, + "learning_rate": 3.183924516579034e-05, + "loss": 0.3045, + "num_input_tokens_seen": 4810016, + "step": 17955 + }, + { + "epoch": 4.707732634338139, + "grad_norm": 1.564977765083313, + "learning_rate": 3.182824347603818e-05, + "loss": 0.2659, + "num_input_tokens_seen": 4811392, + "step": 17960 + }, + { + "epoch": 4.709043250327654, + "grad_norm": 0.9353127479553223, + "learning_rate": 3.181724035714644e-05, + "loss": 0.4178, + "num_input_tokens_seen": 4812944, + "step": 17965 + }, + { + "epoch": 4.710353866317169, + "grad_norm": 0.9428114891052246, + "learning_rate": 3.1806235811418074e-05, + "loss": 0.432, + "num_input_tokens_seen": 4814448, + "step": 17970 + }, + { + "epoch": 4.711664482306684, + "grad_norm": 1.1478612422943115, + "learning_rate": 3.179522984115632e-05, + "loss": 0.3548, + "num_input_tokens_seen": 4815632, + "step": 17975 + }, + { + "epoch": 4.712975098296199, + "grad_norm": 1.5283520221710205, + "learning_rate": 3.178422244866469e-05, + "loss": 0.4047, + "num_input_tokens_seen": 4817120, + "step": 17980 + }, + { + "epoch": 4.714285714285714, + "grad_norm": 0.7798539400100708, + "learning_rate": 3.177321363624703e-05, + "loss": 0.2401, + "num_input_tokens_seen": 4818208, + "step": 17985 + }, + { + "epoch": 4.715596330275229, + "grad_norm": 1.2984323501586914, + "learning_rate": 3.176220340620746e-05, + "loss": 0.2611, + "num_input_tokens_seen": 4819600, + "step": 17990 + }, + { + "epoch": 4.716906946264745, + "grad_norm": 1.1154272556304932, + "learning_rate": 3.17511917608504e-05, + "loss": 0.2894, + "num_input_tokens_seen": 4820672, + "step": 17995 + }, + { + "epoch": 4.718217562254259, + "grad_norm": 1.345892071723938, + "learning_rate": 3.174017870248057e-05, + "loss": 0.3411, + "num_input_tokens_seen": 4821888, + "step": 18000 + }, + { + "epoch": 4.719528178243775, + "grad_norm": 0.7587903738021851, + "learning_rate": 3.172916423340299e-05, + "loss": 0.3261, + "num_input_tokens_seen": 4823504, + "step": 18005 + }, + { + "epoch": 4.7208387942332894, + "grad_norm": 1.181376338005066, + "learning_rate": 3.171814835592294e-05, + "loss": 0.3392, + "num_input_tokens_seen": 4825136, + "step": 18010 + }, + { + "epoch": 4.722149410222805, + "grad_norm": 0.9696865081787109, + "learning_rate": 3.170713107234607e-05, + "loss": 0.5669, + "num_input_tokens_seen": 4826672, + "step": 18015 + }, + { + "epoch": 4.72346002621232, + "grad_norm": 3.3745779991149902, + "learning_rate": 3.169611238497825e-05, + "loss": 0.2781, + "num_input_tokens_seen": 4828496, + "step": 18020 + }, + { + "epoch": 4.724770642201835, + "grad_norm": 1.230749487876892, + "learning_rate": 3.168509229612568e-05, + "loss": 0.5109, + "num_input_tokens_seen": 4829792, + "step": 18025 + }, + { + "epoch": 4.72608125819135, + "grad_norm": 0.7460651993751526, + "learning_rate": 3.167407080809484e-05, + "loss": 0.3996, + "num_input_tokens_seen": 4831216, + "step": 18030 + }, + { + "epoch": 4.727391874180865, + "grad_norm": 1.5483616590499878, + "learning_rate": 3.166304792319251e-05, + "loss": 0.2152, + "num_input_tokens_seen": 4832208, + "step": 18035 + }, + { + "epoch": 4.72870249017038, + "grad_norm": 4.325371742248535, + "learning_rate": 3.165202364372577e-05, + "loss": 0.3311, + "num_input_tokens_seen": 4833616, + "step": 18040 + }, + { + "epoch": 4.730013106159895, + "grad_norm": 0.33442357182502747, + "learning_rate": 3.164099797200197e-05, + "loss": 0.3487, + "num_input_tokens_seen": 4834928, + "step": 18045 + }, + { + "epoch": 4.73132372214941, + "grad_norm": 1.2460216283798218, + "learning_rate": 3.162997091032876e-05, + "loss": 0.5269, + "num_input_tokens_seen": 4836128, + "step": 18050 + }, + { + "epoch": 4.732634338138926, + "grad_norm": 0.9052799344062805, + "learning_rate": 3.161894246101411e-05, + "loss": 0.3095, + "num_input_tokens_seen": 4837504, + "step": 18055 + }, + { + "epoch": 4.73394495412844, + "grad_norm": 1.143214464187622, + "learning_rate": 3.160791262636624e-05, + "loss": 0.3948, + "num_input_tokens_seen": 4838720, + "step": 18060 + }, + { + "epoch": 4.735255570117955, + "grad_norm": 1.352759838104248, + "learning_rate": 3.159688140869366e-05, + "loss": 0.3831, + "num_input_tokens_seen": 4840336, + "step": 18065 + }, + { + "epoch": 4.736566186107471, + "grad_norm": 1.3055419921875, + "learning_rate": 3.158584881030521e-05, + "loss": 0.3427, + "num_input_tokens_seen": 4841424, + "step": 18070 + }, + { + "epoch": 4.737876802096985, + "grad_norm": 0.9903269410133362, + "learning_rate": 3.157481483350998e-05, + "loss": 0.489, + "num_input_tokens_seen": 4842704, + "step": 18075 + }, + { + "epoch": 4.739187418086501, + "grad_norm": 1.083845853805542, + "learning_rate": 3.156377948061737e-05, + "loss": 0.4666, + "num_input_tokens_seen": 4843888, + "step": 18080 + }, + { + "epoch": 4.740498034076015, + "grad_norm": 1.1985441446304321, + "learning_rate": 3.155274275393705e-05, + "loss": 0.2482, + "num_input_tokens_seen": 4844960, + "step": 18085 + }, + { + "epoch": 4.741808650065531, + "grad_norm": 0.6819832921028137, + "learning_rate": 3.1541704655779e-05, + "loss": 0.374, + "num_input_tokens_seen": 4846272, + "step": 18090 + }, + { + "epoch": 4.743119266055046, + "grad_norm": 0.8131844401359558, + "learning_rate": 3.1530665188453464e-05, + "loss": 0.491, + "num_input_tokens_seen": 4847456, + "step": 18095 + }, + { + "epoch": 4.744429882044561, + "grad_norm": 2.2997758388519287, + "learning_rate": 3.151962435427099e-05, + "loss": 0.3397, + "num_input_tokens_seen": 4848560, + "step": 18100 + }, + { + "epoch": 4.745740498034076, + "grad_norm": 0.7484158873558044, + "learning_rate": 3.15085821555424e-05, + "loss": 0.3121, + "num_input_tokens_seen": 4850160, + "step": 18105 + }, + { + "epoch": 4.747051114023591, + "grad_norm": 1.1152859926223755, + "learning_rate": 3.1497538594578814e-05, + "loss": 0.3551, + "num_input_tokens_seen": 4851696, + "step": 18110 + }, + { + "epoch": 4.748361730013106, + "grad_norm": 1.1100409030914307, + "learning_rate": 3.1486493673691624e-05, + "loss": 0.3689, + "num_input_tokens_seen": 4853008, + "step": 18115 + }, + { + "epoch": 4.749672346002622, + "grad_norm": 0.8249077796936035, + "learning_rate": 3.147544739519253e-05, + "loss": 0.4082, + "num_input_tokens_seen": 4854352, + "step": 18120 + }, + { + "epoch": 4.750982961992136, + "grad_norm": 0.6919863224029541, + "learning_rate": 3.1464399761393465e-05, + "loss": 0.2625, + "num_input_tokens_seen": 4855904, + "step": 18125 + }, + { + "epoch": 4.752293577981652, + "grad_norm": 0.8991656303405762, + "learning_rate": 3.145335077460671e-05, + "loss": 0.341, + "num_input_tokens_seen": 4857536, + "step": 18130 + }, + { + "epoch": 4.753604193971166, + "grad_norm": 1.0977232456207275, + "learning_rate": 3.144230043714479e-05, + "loss": 0.2493, + "num_input_tokens_seen": 4858480, + "step": 18135 + }, + { + "epoch": 4.754914809960681, + "grad_norm": 0.5692166090011597, + "learning_rate": 3.143124875132052e-05, + "loss": 0.243, + "num_input_tokens_seen": 4859664, + "step": 18140 + }, + { + "epoch": 4.756225425950197, + "grad_norm": 0.6365553140640259, + "learning_rate": 3.142019571944699e-05, + "loss": 0.2937, + "num_input_tokens_seen": 4861472, + "step": 18145 + }, + { + "epoch": 4.757536041939712, + "grad_norm": 4.606222152709961, + "learning_rate": 3.140914134383759e-05, + "loss": 0.4404, + "num_input_tokens_seen": 4862592, + "step": 18150 + }, + { + "epoch": 4.758846657929227, + "grad_norm": 1.013773798942566, + "learning_rate": 3.139808562680599e-05, + "loss": 0.5672, + "num_input_tokens_seen": 4863696, + "step": 18155 + }, + { + "epoch": 4.760157273918741, + "grad_norm": 1.110933780670166, + "learning_rate": 3.138702857066612e-05, + "loss": 0.2841, + "num_input_tokens_seen": 4864768, + "step": 18160 + }, + { + "epoch": 4.761467889908257, + "grad_norm": 0.6544603109359741, + "learning_rate": 3.137597017773221e-05, + "loss": 0.3204, + "num_input_tokens_seen": 4866112, + "step": 18165 + }, + { + "epoch": 4.762778505897772, + "grad_norm": 0.803669810295105, + "learning_rate": 3.136491045031875e-05, + "loss": 0.2606, + "num_input_tokens_seen": 4867456, + "step": 18170 + }, + { + "epoch": 4.764089121887287, + "grad_norm": 7.43604850769043, + "learning_rate": 3.135384939074054e-05, + "loss": 0.3781, + "num_input_tokens_seen": 4868592, + "step": 18175 + }, + { + "epoch": 4.765399737876802, + "grad_norm": 5.792081356048584, + "learning_rate": 3.1342787001312615e-05, + "loss": 0.406, + "num_input_tokens_seen": 4869824, + "step": 18180 + }, + { + "epoch": 4.766710353866317, + "grad_norm": 0.9842743873596191, + "learning_rate": 3.133172328435034e-05, + "loss": 0.3229, + "num_input_tokens_seen": 4871200, + "step": 18185 + }, + { + "epoch": 4.768020969855832, + "grad_norm": 1.459099292755127, + "learning_rate": 3.132065824216932e-05, + "loss": 0.3253, + "num_input_tokens_seen": 4872224, + "step": 18190 + }, + { + "epoch": 4.7693315858453476, + "grad_norm": 1.0115246772766113, + "learning_rate": 3.1309591877085464e-05, + "loss": 0.557, + "num_input_tokens_seen": 4873520, + "step": 18195 + }, + { + "epoch": 4.770642201834862, + "grad_norm": 1.2025846242904663, + "learning_rate": 3.1298524191414915e-05, + "loss": 0.3565, + "num_input_tokens_seen": 4874992, + "step": 18200 + }, + { + "epoch": 4.771952817824378, + "grad_norm": 1.0505247116088867, + "learning_rate": 3.128745518747414e-05, + "loss": 0.2462, + "num_input_tokens_seen": 4876112, + "step": 18205 + }, + { + "epoch": 4.773263433813892, + "grad_norm": 1.4002015590667725, + "learning_rate": 3.127638486757985e-05, + "loss": 0.1992, + "num_input_tokens_seen": 4877136, + "step": 18210 + }, + { + "epoch": 4.774574049803408, + "grad_norm": 0.7461525797843933, + "learning_rate": 3.126531323404907e-05, + "loss": 0.2401, + "num_input_tokens_seen": 4878848, + "step": 18215 + }, + { + "epoch": 4.775884665792923, + "grad_norm": 0.7677587270736694, + "learning_rate": 3.125424028919904e-05, + "loss": 0.2499, + "num_input_tokens_seen": 4880912, + "step": 18220 + }, + { + "epoch": 4.777195281782438, + "grad_norm": 1.0932437181472778, + "learning_rate": 3.1243166035347335e-05, + "loss": 0.3984, + "num_input_tokens_seen": 4882448, + "step": 18225 + }, + { + "epoch": 4.778505897771953, + "grad_norm": 1.1859028339385986, + "learning_rate": 3.123209047481177e-05, + "loss": 0.5212, + "num_input_tokens_seen": 4883792, + "step": 18230 + }, + { + "epoch": 4.779816513761467, + "grad_norm": 2.1577224731445312, + "learning_rate": 3.1221013609910424e-05, + "loss": 0.4415, + "num_input_tokens_seen": 4884928, + "step": 18235 + }, + { + "epoch": 4.781127129750983, + "grad_norm": 1.1093201637268066, + "learning_rate": 3.120993544296168e-05, + "loss": 0.4318, + "num_input_tokens_seen": 4886288, + "step": 18240 + }, + { + "epoch": 4.7824377457404985, + "grad_norm": 0.9409151673316956, + "learning_rate": 3.119885597628418e-05, + "loss": 0.3677, + "num_input_tokens_seen": 4887712, + "step": 18245 + }, + { + "epoch": 4.783748361730013, + "grad_norm": 0.9634373188018799, + "learning_rate": 3.1187775212196835e-05, + "loss": 0.2729, + "num_input_tokens_seen": 4888880, + "step": 18250 + }, + { + "epoch": 4.785058977719528, + "grad_norm": 0.8128646612167358, + "learning_rate": 3.117669315301883e-05, + "loss": 0.4057, + "num_input_tokens_seen": 4890208, + "step": 18255 + }, + { + "epoch": 4.786369593709043, + "grad_norm": 1.173899531364441, + "learning_rate": 3.116560980106962e-05, + "loss": 0.2814, + "num_input_tokens_seen": 4891440, + "step": 18260 + }, + { + "epoch": 4.787680209698558, + "grad_norm": 0.9088613390922546, + "learning_rate": 3.115452515866893e-05, + "loss": 0.3121, + "num_input_tokens_seen": 4892592, + "step": 18265 + }, + { + "epoch": 4.7889908256880735, + "grad_norm": 0.7013163566589355, + "learning_rate": 3.114343922813677e-05, + "loss": 0.2922, + "num_input_tokens_seen": 4893856, + "step": 18270 + }, + { + "epoch": 4.790301441677588, + "grad_norm": 0.7342329621315002, + "learning_rate": 3.1132352011793384e-05, + "loss": 0.3539, + "num_input_tokens_seen": 4895120, + "step": 18275 + }, + { + "epoch": 4.791612057667104, + "grad_norm": 1.3154577016830444, + "learning_rate": 3.112126351195932e-05, + "loss": 0.3062, + "num_input_tokens_seen": 4896512, + "step": 18280 + }, + { + "epoch": 4.792922673656618, + "grad_norm": 1.0527985095977783, + "learning_rate": 3.1110173730955374e-05, + "loss": 0.4392, + "num_input_tokens_seen": 4897824, + "step": 18285 + }, + { + "epoch": 4.794233289646134, + "grad_norm": 1.133874773979187, + "learning_rate": 3.109908267110263e-05, + "loss": 0.4336, + "num_input_tokens_seen": 4899184, + "step": 18290 + }, + { + "epoch": 4.795543905635649, + "grad_norm": 0.9508861303329468, + "learning_rate": 3.108799033472242e-05, + "loss": 0.3874, + "num_input_tokens_seen": 4900976, + "step": 18295 + }, + { + "epoch": 4.796854521625164, + "grad_norm": 1.1436272859573364, + "learning_rate": 3.107689672413636e-05, + "loss": 0.2782, + "num_input_tokens_seen": 4902480, + "step": 18300 + }, + { + "epoch": 4.798165137614679, + "grad_norm": 0.6804212331771851, + "learning_rate": 3.10658018416663e-05, + "loss": 0.2393, + "num_input_tokens_seen": 4903744, + "step": 18305 + }, + { + "epoch": 4.799475753604194, + "grad_norm": 0.7254925966262817, + "learning_rate": 3.1054705689634405e-05, + "loss": 0.2647, + "num_input_tokens_seen": 4905216, + "step": 18310 + }, + { + "epoch": 4.800786369593709, + "grad_norm": 0.9516083598136902, + "learning_rate": 3.1043608270363075e-05, + "loss": 0.3005, + "num_input_tokens_seen": 4906512, + "step": 18315 + }, + { + "epoch": 4.8020969855832245, + "grad_norm": 0.5886860489845276, + "learning_rate": 3.103250958617497e-05, + "loss": 0.3023, + "num_input_tokens_seen": 4907808, + "step": 18320 + }, + { + "epoch": 4.803407601572739, + "grad_norm": 0.7668098211288452, + "learning_rate": 3.102140963939303e-05, + "loss": 0.4699, + "num_input_tokens_seen": 4909728, + "step": 18325 + }, + { + "epoch": 4.804718217562254, + "grad_norm": 1.0779989957809448, + "learning_rate": 3.101030843234046e-05, + "loss": 0.3518, + "num_input_tokens_seen": 4910896, + "step": 18330 + }, + { + "epoch": 4.806028833551769, + "grad_norm": 1.1137720346450806, + "learning_rate": 3.099920596734073e-05, + "loss": 0.5141, + "num_input_tokens_seen": 4912288, + "step": 18335 + }, + { + "epoch": 4.807339449541285, + "grad_norm": 0.9120848774909973, + "learning_rate": 3.0988102246717544e-05, + "loss": 0.2021, + "num_input_tokens_seen": 4913312, + "step": 18340 + }, + { + "epoch": 4.8086500655307995, + "grad_norm": 2.549769163131714, + "learning_rate": 3.0976997272794906e-05, + "loss": 0.2819, + "num_input_tokens_seen": 4914544, + "step": 18345 + }, + { + "epoch": 4.809960681520314, + "grad_norm": 1.3887373208999634, + "learning_rate": 3.096589104789707e-05, + "loss": 0.3158, + "num_input_tokens_seen": 4915776, + "step": 18350 + }, + { + "epoch": 4.81127129750983, + "grad_norm": 1.4607259035110474, + "learning_rate": 3.095478357434854e-05, + "loss": 0.481, + "num_input_tokens_seen": 4917008, + "step": 18355 + }, + { + "epoch": 4.812581913499344, + "grad_norm": 2.8391811847686768, + "learning_rate": 3.09436748544741e-05, + "loss": 0.2442, + "num_input_tokens_seen": 4918032, + "step": 18360 + }, + { + "epoch": 4.81389252948886, + "grad_norm": 0.9152385592460632, + "learning_rate": 3.093256489059878e-05, + "loss": 0.3597, + "num_input_tokens_seen": 4919568, + "step": 18365 + }, + { + "epoch": 4.815203145478375, + "grad_norm": 1.5025928020477295, + "learning_rate": 3.092145368504789e-05, + "loss": 0.3711, + "num_input_tokens_seen": 4920912, + "step": 18370 + }, + { + "epoch": 4.81651376146789, + "grad_norm": 1.2922313213348389, + "learning_rate": 3.0910341240146966e-05, + "loss": 0.3503, + "num_input_tokens_seen": 4922160, + "step": 18375 + }, + { + "epoch": 4.817824377457405, + "grad_norm": 1.75254225730896, + "learning_rate": 3.089922755822183e-05, + "loss": 0.4129, + "num_input_tokens_seen": 4923472, + "step": 18380 + }, + { + "epoch": 4.81913499344692, + "grad_norm": 0.7870863676071167, + "learning_rate": 3.088811264159856e-05, + "loss": 0.4643, + "num_input_tokens_seen": 4924528, + "step": 18385 + }, + { + "epoch": 4.820445609436435, + "grad_norm": 1.075412631034851, + "learning_rate": 3.08769964926035e-05, + "loss": 0.3075, + "num_input_tokens_seen": 4925936, + "step": 18390 + }, + { + "epoch": 4.8217562254259505, + "grad_norm": 1.3078413009643555, + "learning_rate": 3.086587911356322e-05, + "loss": 0.4863, + "num_input_tokens_seen": 4927504, + "step": 18395 + }, + { + "epoch": 4.823066841415465, + "grad_norm": 1.105099081993103, + "learning_rate": 3.085476050680458e-05, + "loss": 0.3834, + "num_input_tokens_seen": 4928800, + "step": 18400 + }, + { + "epoch": 4.824377457404981, + "grad_norm": 1.3392224311828613, + "learning_rate": 3.0843640674654675e-05, + "loss": 0.3166, + "num_input_tokens_seen": 4930480, + "step": 18405 + }, + { + "epoch": 4.825688073394495, + "grad_norm": 0.5617567300796509, + "learning_rate": 3.083251961944088e-05, + "loss": 0.3434, + "num_input_tokens_seen": 4931904, + "step": 18410 + }, + { + "epoch": 4.826998689384011, + "grad_norm": 1.043297529220581, + "learning_rate": 3.08213973434908e-05, + "loss": 0.3121, + "num_input_tokens_seen": 4933728, + "step": 18415 + }, + { + "epoch": 4.8283093053735255, + "grad_norm": 1.0914726257324219, + "learning_rate": 3.081027384913232e-05, + "loss": 0.486, + "num_input_tokens_seen": 4935264, + "step": 18420 + }, + { + "epoch": 4.82961992136304, + "grad_norm": 2.1329591274261475, + "learning_rate": 3.079914913869356e-05, + "loss": 0.415, + "num_input_tokens_seen": 4936736, + "step": 18425 + }, + { + "epoch": 4.830930537352556, + "grad_norm": 1.3426132202148438, + "learning_rate": 3.0788023214502915e-05, + "loss": 0.3625, + "num_input_tokens_seen": 4938448, + "step": 18430 + }, + { + "epoch": 4.832241153342071, + "grad_norm": 1.8183039426803589, + "learning_rate": 3.0776896078889e-05, + "loss": 0.304, + "num_input_tokens_seen": 4939632, + "step": 18435 + }, + { + "epoch": 4.833551769331586, + "grad_norm": 3.670880079269409, + "learning_rate": 3.0765767734180714e-05, + "loss": 0.3283, + "num_input_tokens_seen": 4940800, + "step": 18440 + }, + { + "epoch": 4.834862385321101, + "grad_norm": 0.807392418384552, + "learning_rate": 3.075463818270722e-05, + "loss": 0.3344, + "num_input_tokens_seen": 4942032, + "step": 18445 + }, + { + "epoch": 4.836173001310616, + "grad_norm": 0.23548486828804016, + "learning_rate": 3.074350742679789e-05, + "loss": 0.2534, + "num_input_tokens_seen": 4943312, + "step": 18450 + }, + { + "epoch": 4.837483617300131, + "grad_norm": 2.3639309406280518, + "learning_rate": 3.0732375468782384e-05, + "loss": 0.4618, + "num_input_tokens_seen": 4944448, + "step": 18455 + }, + { + "epoch": 4.838794233289646, + "grad_norm": 0.8570098280906677, + "learning_rate": 3.072124231099058e-05, + "loss": 0.2838, + "num_input_tokens_seen": 4946000, + "step": 18460 + }, + { + "epoch": 4.840104849279161, + "grad_norm": 0.9788373708724976, + "learning_rate": 3.0710107955752664e-05, + "loss": 0.3404, + "num_input_tokens_seen": 4947152, + "step": 18465 + }, + { + "epoch": 4.8414154652686765, + "grad_norm": 0.6536195278167725, + "learning_rate": 3.069897240539901e-05, + "loss": 0.3137, + "num_input_tokens_seen": 4948704, + "step": 18470 + }, + { + "epoch": 4.842726081258191, + "grad_norm": 0.7625897526741028, + "learning_rate": 3.068783566226028e-05, + "loss": 0.3581, + "num_input_tokens_seen": 4950096, + "step": 18475 + }, + { + "epoch": 4.844036697247707, + "grad_norm": 0.9314426779747009, + "learning_rate": 3.067669772866738e-05, + "loss": 0.4159, + "num_input_tokens_seen": 4951408, + "step": 18480 + }, + { + "epoch": 4.845347313237221, + "grad_norm": 3.504629135131836, + "learning_rate": 3.066555860695145e-05, + "loss": 0.5554, + "num_input_tokens_seen": 4952352, + "step": 18485 + }, + { + "epoch": 4.846657929226737, + "grad_norm": 2.742645025253296, + "learning_rate": 3.065441829944389e-05, + "loss": 0.4769, + "num_input_tokens_seen": 4953696, + "step": 18490 + }, + { + "epoch": 4.8479685452162515, + "grad_norm": 0.5886386632919312, + "learning_rate": 3.064327680847635e-05, + "loss": 0.2841, + "num_input_tokens_seen": 4955376, + "step": 18495 + }, + { + "epoch": 4.849279161205767, + "grad_norm": 1.2254918813705444, + "learning_rate": 3.063213413638073e-05, + "loss": 0.3935, + "num_input_tokens_seen": 4957024, + "step": 18500 + }, + { + "epoch": 4.850589777195282, + "grad_norm": 0.8790766596794128, + "learning_rate": 3.0620990285489165e-05, + "loss": 0.4453, + "num_input_tokens_seen": 4958272, + "step": 18505 + }, + { + "epoch": 4.851900393184797, + "grad_norm": 1.1389793157577515, + "learning_rate": 3.0609845258134046e-05, + "loss": 0.391, + "num_input_tokens_seen": 4959328, + "step": 18510 + }, + { + "epoch": 4.853211009174312, + "grad_norm": 1.5086239576339722, + "learning_rate": 3.0598699056648005e-05, + "loss": 0.3196, + "num_input_tokens_seen": 4960352, + "step": 18515 + }, + { + "epoch": 4.854521625163827, + "grad_norm": 1.9159107208251953, + "learning_rate": 3.0587551683363926e-05, + "loss": 0.3225, + "num_input_tokens_seen": 4961488, + "step": 18520 + }, + { + "epoch": 4.855832241153342, + "grad_norm": 1.257799744606018, + "learning_rate": 3.057640314061493e-05, + "loss": 0.3457, + "num_input_tokens_seen": 4962816, + "step": 18525 + }, + { + "epoch": 4.857142857142857, + "grad_norm": 0.5960150957107544, + "learning_rate": 3.05652534307344e-05, + "loss": 0.2444, + "num_input_tokens_seen": 4964256, + "step": 18530 + }, + { + "epoch": 4.858453473132372, + "grad_norm": 0.1042783111333847, + "learning_rate": 3.055410255605594e-05, + "loss": 0.3234, + "num_input_tokens_seen": 4965680, + "step": 18535 + }, + { + "epoch": 4.859764089121887, + "grad_norm": 3.23889422416687, + "learning_rate": 3.054295051891341e-05, + "loss": 0.3004, + "num_input_tokens_seen": 4967216, + "step": 18540 + }, + { + "epoch": 4.8610747051114025, + "grad_norm": 1.5551230907440186, + "learning_rate": 3.053179732164092e-05, + "loss": 0.2402, + "num_input_tokens_seen": 4968496, + "step": 18545 + }, + { + "epoch": 4.862385321100917, + "grad_norm": 4.364006996154785, + "learning_rate": 3.0520642966572805e-05, + "loss": 0.4042, + "num_input_tokens_seen": 4970112, + "step": 18550 + }, + { + "epoch": 4.863695937090433, + "grad_norm": 1.3413207530975342, + "learning_rate": 3.050948745604365e-05, + "loss": 0.3714, + "num_input_tokens_seen": 4971568, + "step": 18555 + }, + { + "epoch": 4.865006553079947, + "grad_norm": 2.2098522186279297, + "learning_rate": 3.049833079238829e-05, + "loss": 0.6908, + "num_input_tokens_seen": 4972704, + "step": 18560 + }, + { + "epoch": 4.866317169069463, + "grad_norm": 1.4211499691009521, + "learning_rate": 3.0487172977941806e-05, + "loss": 0.307, + "num_input_tokens_seen": 4973936, + "step": 18565 + }, + { + "epoch": 4.8676277850589775, + "grad_norm": 1.444441556930542, + "learning_rate": 3.0476014015039496e-05, + "loss": 0.3311, + "num_input_tokens_seen": 4975184, + "step": 18570 + }, + { + "epoch": 4.868938401048493, + "grad_norm": 1.1030058860778809, + "learning_rate": 3.046485390601691e-05, + "loss": 0.294, + "num_input_tokens_seen": 4976304, + "step": 18575 + }, + { + "epoch": 4.870249017038008, + "grad_norm": 0.6408756971359253, + "learning_rate": 3.045369265320984e-05, + "loss": 0.3968, + "num_input_tokens_seen": 4977696, + "step": 18580 + }, + { + "epoch": 4.871559633027523, + "grad_norm": 0.8713457584381104, + "learning_rate": 3.0442530258954334e-05, + "loss": 0.4869, + "num_input_tokens_seen": 4979360, + "step": 18585 + }, + { + "epoch": 4.872870249017038, + "grad_norm": 0.8469298481941223, + "learning_rate": 3.0431366725586642e-05, + "loss": 0.3285, + "num_input_tokens_seen": 4980656, + "step": 18590 + }, + { + "epoch": 4.8741808650065535, + "grad_norm": 0.7236356735229492, + "learning_rate": 3.0420202055443276e-05, + "loss": 0.2647, + "num_input_tokens_seen": 4981920, + "step": 18595 + }, + { + "epoch": 4.875491480996068, + "grad_norm": 2.0115246772766113, + "learning_rate": 3.040903625086099e-05, + "loss": 0.3432, + "num_input_tokens_seen": 4983136, + "step": 18600 + }, + { + "epoch": 4.876802096985584, + "grad_norm": 0.9456266164779663, + "learning_rate": 3.0397869314176762e-05, + "loss": 0.4865, + "num_input_tokens_seen": 4984304, + "step": 18605 + }, + { + "epoch": 4.878112712975098, + "grad_norm": 0.7484197020530701, + "learning_rate": 3.0386701247727807e-05, + "loss": 0.357, + "num_input_tokens_seen": 4985824, + "step": 18610 + }, + { + "epoch": 4.879423328964613, + "grad_norm": 1.5235917568206787, + "learning_rate": 3.0375532053851592e-05, + "loss": 0.43, + "num_input_tokens_seen": 4987280, + "step": 18615 + }, + { + "epoch": 4.8807339449541285, + "grad_norm": 0.75788813829422, + "learning_rate": 3.036436173488581e-05, + "loss": 0.215, + "num_input_tokens_seen": 4988272, + "step": 18620 + }, + { + "epoch": 4.882044560943643, + "grad_norm": 1.6043989658355713, + "learning_rate": 3.0353190293168388e-05, + "loss": 0.2061, + "num_input_tokens_seen": 4989392, + "step": 18625 + }, + { + "epoch": 4.883355176933159, + "grad_norm": 0.6732480525970459, + "learning_rate": 3.0342017731037482e-05, + "loss": 0.3957, + "num_input_tokens_seen": 4991296, + "step": 18630 + }, + { + "epoch": 4.884665792922673, + "grad_norm": 0.792560338973999, + "learning_rate": 3.0330844050831492e-05, + "loss": 0.3974, + "num_input_tokens_seen": 4992720, + "step": 18635 + }, + { + "epoch": 4.885976408912189, + "grad_norm": 1.254335880279541, + "learning_rate": 3.0319669254889055e-05, + "loss": 0.3904, + "num_input_tokens_seen": 4993856, + "step": 18640 + }, + { + "epoch": 4.8872870249017035, + "grad_norm": 0.9661917090415955, + "learning_rate": 3.030849334554904e-05, + "loss": 0.3117, + "num_input_tokens_seen": 4994864, + "step": 18645 + }, + { + "epoch": 4.888597640891219, + "grad_norm": 1.5746444463729858, + "learning_rate": 3.0297316325150537e-05, + "loss": 0.3014, + "num_input_tokens_seen": 4996048, + "step": 18650 + }, + { + "epoch": 4.889908256880734, + "grad_norm": 1.3126163482666016, + "learning_rate": 3.0286138196032876e-05, + "loss": 0.3714, + "num_input_tokens_seen": 4997568, + "step": 18655 + }, + { + "epoch": 4.891218872870249, + "grad_norm": 1.3633798360824585, + "learning_rate": 3.0274958960535632e-05, + "loss": 0.3924, + "num_input_tokens_seen": 4998640, + "step": 18660 + }, + { + "epoch": 4.892529488859764, + "grad_norm": 1.8840200901031494, + "learning_rate": 3.0263778620998583e-05, + "loss": 0.4742, + "num_input_tokens_seen": 5000064, + "step": 18665 + }, + { + "epoch": 4.8938401048492794, + "grad_norm": 1.0625131130218506, + "learning_rate": 3.025259717976177e-05, + "loss": 0.1828, + "num_input_tokens_seen": 5001152, + "step": 18670 + }, + { + "epoch": 4.895150720838794, + "grad_norm": 0.822849690914154, + "learning_rate": 3.0241414639165444e-05, + "loss": 0.3156, + "num_input_tokens_seen": 5002480, + "step": 18675 + }, + { + "epoch": 4.89646133682831, + "grad_norm": 2.9063117504119873, + "learning_rate": 3.0230231001550086e-05, + "loss": 0.3453, + "num_input_tokens_seen": 5003696, + "step": 18680 + }, + { + "epoch": 4.897771952817824, + "grad_norm": 0.6944376826286316, + "learning_rate": 3.0219046269256425e-05, + "loss": 0.2739, + "num_input_tokens_seen": 5005312, + "step": 18685 + }, + { + "epoch": 4.89908256880734, + "grad_norm": 7.326805114746094, + "learning_rate": 3.020786044462539e-05, + "loss": 0.4178, + "num_input_tokens_seen": 5006688, + "step": 18690 + }, + { + "epoch": 4.9003931847968545, + "grad_norm": 0.918396532535553, + "learning_rate": 3.019667352999816e-05, + "loss": 0.4, + "num_input_tokens_seen": 5008160, + "step": 18695 + }, + { + "epoch": 4.90170380078637, + "grad_norm": 1.788483738899231, + "learning_rate": 3.018548552771615e-05, + "loss": 0.4515, + "num_input_tokens_seen": 5009744, + "step": 18700 + }, + { + "epoch": 4.903014416775885, + "grad_norm": 0.5679213404655457, + "learning_rate": 3.0174296440120968e-05, + "loss": 0.2791, + "num_input_tokens_seen": 5010944, + "step": 18705 + }, + { + "epoch": 4.904325032765399, + "grad_norm": 0.7186155319213867, + "learning_rate": 3.0163106269554485e-05, + "loss": 0.4343, + "num_input_tokens_seen": 5012816, + "step": 18710 + }, + { + "epoch": 4.905635648754915, + "grad_norm": 1.1662487983703613, + "learning_rate": 3.0151915018358785e-05, + "loss": 0.3745, + "num_input_tokens_seen": 5013936, + "step": 18715 + }, + { + "epoch": 4.9069462647444295, + "grad_norm": 1.0467970371246338, + "learning_rate": 3.0140722688876166e-05, + "loss": 0.3716, + "num_input_tokens_seen": 5015344, + "step": 18720 + }, + { + "epoch": 4.908256880733945, + "grad_norm": 0.5466160774230957, + "learning_rate": 3.0129529283449177e-05, + "loss": 0.2696, + "num_input_tokens_seen": 5016544, + "step": 18725 + }, + { + "epoch": 4.90956749672346, + "grad_norm": 0.9725759029388428, + "learning_rate": 3.011833480442057e-05, + "loss": 0.362, + "num_input_tokens_seen": 5018128, + "step": 18730 + }, + { + "epoch": 4.910878112712975, + "grad_norm": 1.0327916145324707, + "learning_rate": 3.0107139254133332e-05, + "loss": 0.3926, + "num_input_tokens_seen": 5019504, + "step": 18735 + }, + { + "epoch": 4.91218872870249, + "grad_norm": 1.0019896030426025, + "learning_rate": 3.009594263493068e-05, + "loss": 0.2918, + "num_input_tokens_seen": 5021040, + "step": 18740 + }, + { + "epoch": 4.913499344692005, + "grad_norm": 3.1465892791748047, + "learning_rate": 3.008474494915603e-05, + "loss": 0.3305, + "num_input_tokens_seen": 5022304, + "step": 18745 + }, + { + "epoch": 4.91480996068152, + "grad_norm": 1.6420344114303589, + "learning_rate": 3.0073546199153053e-05, + "loss": 0.3381, + "num_input_tokens_seen": 5023808, + "step": 18750 + }, + { + "epoch": 4.916120576671036, + "grad_norm": 1.3414525985717773, + "learning_rate": 3.006234638726561e-05, + "loss": 0.4585, + "num_input_tokens_seen": 5025200, + "step": 18755 + }, + { + "epoch": 4.91743119266055, + "grad_norm": 1.4028342962265015, + "learning_rate": 3.005114551583783e-05, + "loss": 0.3228, + "num_input_tokens_seen": 5026656, + "step": 18760 + }, + { + "epoch": 4.918741808650066, + "grad_norm": 0.9323918223381042, + "learning_rate": 3.003994358721402e-05, + "loss": 0.2871, + "num_input_tokens_seen": 5027616, + "step": 18765 + }, + { + "epoch": 4.9200524246395805, + "grad_norm": 5.245394706726074, + "learning_rate": 3.002874060373872e-05, + "loss": 0.6214, + "num_input_tokens_seen": 5029584, + "step": 18770 + }, + { + "epoch": 4.921363040629096, + "grad_norm": 0.76417076587677, + "learning_rate": 3.0017536567756705e-05, + "loss": 0.2996, + "num_input_tokens_seen": 5030672, + "step": 18775 + }, + { + "epoch": 4.922673656618611, + "grad_norm": 0.6253792643547058, + "learning_rate": 3.0006331481612955e-05, + "loss": 0.289, + "num_input_tokens_seen": 5032352, + "step": 18780 + }, + { + "epoch": 4.923984272608125, + "grad_norm": 0.9163411855697632, + "learning_rate": 2.999512534765267e-05, + "loss": 0.1844, + "num_input_tokens_seen": 5033744, + "step": 18785 + }, + { + "epoch": 4.925294888597641, + "grad_norm": 1.120749592781067, + "learning_rate": 2.9983918168221287e-05, + "loss": 0.3214, + "num_input_tokens_seen": 5034800, + "step": 18790 + }, + { + "epoch": 4.926605504587156, + "grad_norm": 0.9863185882568359, + "learning_rate": 2.997270994566444e-05, + "loss": 0.3791, + "num_input_tokens_seen": 5035952, + "step": 18795 + }, + { + "epoch": 4.927916120576671, + "grad_norm": 0.6060236096382141, + "learning_rate": 2.9961500682328e-05, + "loss": 0.2657, + "num_input_tokens_seen": 5037088, + "step": 18800 + }, + { + "epoch": 4.929226736566186, + "grad_norm": 0.9339101910591125, + "learning_rate": 2.995029038055804e-05, + "loss": 0.6195, + "num_input_tokens_seen": 5038624, + "step": 18805 + }, + { + "epoch": 4.930537352555701, + "grad_norm": 0.9192453026771545, + "learning_rate": 2.9939079042700847e-05, + "loss": 0.3988, + "num_input_tokens_seen": 5040416, + "step": 18810 + }, + { + "epoch": 4.931847968545216, + "grad_norm": 1.1175365447998047, + "learning_rate": 2.9927866671102946e-05, + "loss": 0.332, + "num_input_tokens_seen": 5041696, + "step": 18815 + }, + { + "epoch": 4.933158584534731, + "grad_norm": 1.5077769756317139, + "learning_rate": 2.9916653268111074e-05, + "loss": 0.2579, + "num_input_tokens_seen": 5043168, + "step": 18820 + }, + { + "epoch": 4.934469200524246, + "grad_norm": 0.6772891879081726, + "learning_rate": 2.9905438836072165e-05, + "loss": 0.4346, + "num_input_tokens_seen": 5044736, + "step": 18825 + }, + { + "epoch": 4.935779816513762, + "grad_norm": 1.2520155906677246, + "learning_rate": 2.989422337733338e-05, + "loss": 0.4328, + "num_input_tokens_seen": 5046256, + "step": 18830 + }, + { + "epoch": 4.937090432503276, + "grad_norm": 1.3768835067749023, + "learning_rate": 2.98830068942421e-05, + "loss": 0.4867, + "num_input_tokens_seen": 5047536, + "step": 18835 + }, + { + "epoch": 4.938401048492792, + "grad_norm": 1.226650357246399, + "learning_rate": 2.987178938914591e-05, + "loss": 0.5391, + "num_input_tokens_seen": 5048768, + "step": 18840 + }, + { + "epoch": 4.9397116644823065, + "grad_norm": 1.6564264297485352, + "learning_rate": 2.9860570864392625e-05, + "loss": 0.3818, + "num_input_tokens_seen": 5049760, + "step": 18845 + }, + { + "epoch": 4.941022280471822, + "grad_norm": 2.4391252994537354, + "learning_rate": 2.9849351322330255e-05, + "loss": 0.34, + "num_input_tokens_seen": 5050720, + "step": 18850 + }, + { + "epoch": 4.942332896461337, + "grad_norm": 0.671851396560669, + "learning_rate": 2.983813076530703e-05, + "loss": 0.3276, + "num_input_tokens_seen": 5051808, + "step": 18855 + }, + { + "epoch": 4.943643512450852, + "grad_norm": 1.0996750593185425, + "learning_rate": 2.9826909195671403e-05, + "loss": 0.156, + "num_input_tokens_seen": 5053168, + "step": 18860 + }, + { + "epoch": 4.944954128440367, + "grad_norm": 1.154911756515503, + "learning_rate": 2.981568661577202e-05, + "loss": 0.4453, + "num_input_tokens_seen": 5054368, + "step": 18865 + }, + { + "epoch": 4.946264744429882, + "grad_norm": 1.4686685800552368, + "learning_rate": 2.9804463027957736e-05, + "loss": 0.5692, + "num_input_tokens_seen": 5055600, + "step": 18870 + }, + { + "epoch": 4.947575360419397, + "grad_norm": 0.8487752676010132, + "learning_rate": 2.9793238434577652e-05, + "loss": 0.3101, + "num_input_tokens_seen": 5057040, + "step": 18875 + }, + { + "epoch": 4.948885976408912, + "grad_norm": 1.5182559490203857, + "learning_rate": 2.978201283798105e-05, + "loss": 0.2243, + "num_input_tokens_seen": 5059248, + "step": 18880 + }, + { + "epoch": 4.950196592398427, + "grad_norm": 0.7365012764930725, + "learning_rate": 2.977078624051742e-05, + "loss": 0.3407, + "num_input_tokens_seen": 5060640, + "step": 18885 + }, + { + "epoch": 4.951507208387943, + "grad_norm": 1.3879543542861938, + "learning_rate": 2.9759558644536467e-05, + "loss": 0.3933, + "num_input_tokens_seen": 5062032, + "step": 18890 + }, + { + "epoch": 4.952817824377457, + "grad_norm": 1.3860492706298828, + "learning_rate": 2.9748330052388117e-05, + "loss": 0.3363, + "num_input_tokens_seen": 5063424, + "step": 18895 + }, + { + "epoch": 4.954128440366972, + "grad_norm": 1.5580660104751587, + "learning_rate": 2.9737100466422498e-05, + "loss": 0.3541, + "num_input_tokens_seen": 5064384, + "step": 18900 + }, + { + "epoch": 4.955439056356488, + "grad_norm": 1.17000150680542, + "learning_rate": 2.9725869888989926e-05, + "loss": 0.2861, + "num_input_tokens_seen": 5065728, + "step": 18905 + }, + { + "epoch": 4.956749672346002, + "grad_norm": 0.7257784008979797, + "learning_rate": 2.9714638322440956e-05, + "loss": 0.2758, + "num_input_tokens_seen": 5067408, + "step": 18910 + }, + { + "epoch": 4.958060288335518, + "grad_norm": 1.4601056575775146, + "learning_rate": 2.9703405769126334e-05, + "loss": 0.3925, + "num_input_tokens_seen": 5068592, + "step": 18915 + }, + { + "epoch": 4.9593709043250325, + "grad_norm": 1.5336244106292725, + "learning_rate": 2.969217223139702e-05, + "loss": 0.4236, + "num_input_tokens_seen": 5070224, + "step": 18920 + }, + { + "epoch": 4.960681520314548, + "grad_norm": 1.305870532989502, + "learning_rate": 2.9680937711604146e-05, + "loss": 0.3271, + "num_input_tokens_seen": 5071424, + "step": 18925 + }, + { + "epoch": 4.961992136304063, + "grad_norm": 0.8357679843902588, + "learning_rate": 2.9669702212099104e-05, + "loss": 0.2606, + "num_input_tokens_seen": 5072880, + "step": 18930 + }, + { + "epoch": 4.963302752293578, + "grad_norm": 1.022877812385559, + "learning_rate": 2.9658465735233467e-05, + "loss": 0.3071, + "num_input_tokens_seen": 5074128, + "step": 18935 + }, + { + "epoch": 4.964613368283093, + "grad_norm": 1.900709629058838, + "learning_rate": 2.9647228283358995e-05, + "loss": 0.4616, + "num_input_tokens_seen": 5075216, + "step": 18940 + }, + { + "epoch": 4.965923984272608, + "grad_norm": 3.828185558319092, + "learning_rate": 2.9635989858827673e-05, + "loss": 0.5076, + "num_input_tokens_seen": 5076368, + "step": 18945 + }, + { + "epoch": 4.967234600262123, + "grad_norm": 2.133880376815796, + "learning_rate": 2.962475046399168e-05, + "loss": 0.5347, + "num_input_tokens_seen": 5077760, + "step": 18950 + }, + { + "epoch": 4.968545216251639, + "grad_norm": 0.9677647948265076, + "learning_rate": 2.9613510101203406e-05, + "loss": 0.3721, + "num_input_tokens_seen": 5079728, + "step": 18955 + }, + { + "epoch": 4.969855832241153, + "grad_norm": 2.693253755569458, + "learning_rate": 2.9602268772815445e-05, + "loss": 0.3223, + "num_input_tokens_seen": 5081232, + "step": 18960 + }, + { + "epoch": 4.971166448230669, + "grad_norm": 1.5019992589950562, + "learning_rate": 2.959102648118058e-05, + "loss": 0.5195, + "num_input_tokens_seen": 5082672, + "step": 18965 + }, + { + "epoch": 4.972477064220183, + "grad_norm": 1.2759748697280884, + "learning_rate": 2.9579783228651802e-05, + "loss": 0.4022, + "num_input_tokens_seen": 5083824, + "step": 18970 + }, + { + "epoch": 4.973787680209698, + "grad_norm": 0.8989389538764954, + "learning_rate": 2.956853901758231e-05, + "loss": 0.2929, + "num_input_tokens_seen": 5084912, + "step": 18975 + }, + { + "epoch": 4.975098296199214, + "grad_norm": 0.8744938969612122, + "learning_rate": 2.9557293850325496e-05, + "loss": 0.3432, + "num_input_tokens_seen": 5086528, + "step": 18980 + }, + { + "epoch": 4.976408912188729, + "grad_norm": 1.5552090406417847, + "learning_rate": 2.954604772923495e-05, + "loss": 0.2623, + "num_input_tokens_seen": 5087856, + "step": 18985 + }, + { + "epoch": 4.977719528178244, + "grad_norm": 1.2533988952636719, + "learning_rate": 2.9534800656664468e-05, + "loss": 0.3802, + "num_input_tokens_seen": 5089328, + "step": 18990 + }, + { + "epoch": 4.9790301441677585, + "grad_norm": 0.8204338550567627, + "learning_rate": 2.9523552634968056e-05, + "loss": 0.2767, + "num_input_tokens_seen": 5090400, + "step": 18995 + }, + { + "epoch": 4.980340760157274, + "grad_norm": 1.1097732782363892, + "learning_rate": 2.9512303666499886e-05, + "loss": 0.2871, + "num_input_tokens_seen": 5091984, + "step": 19000 + }, + { + "epoch": 4.981651376146789, + "grad_norm": 1.0312480926513672, + "learning_rate": 2.9501053753614354e-05, + "loss": 0.3018, + "num_input_tokens_seen": 5093424, + "step": 19005 + }, + { + "epoch": 4.982961992136304, + "grad_norm": 7.7398152351379395, + "learning_rate": 2.9489802898666042e-05, + "loss": 0.3059, + "num_input_tokens_seen": 5094752, + "step": 19010 + }, + { + "epoch": 4.984272608125819, + "grad_norm": 1.4810054302215576, + "learning_rate": 2.9478551104009756e-05, + "loss": 0.3256, + "num_input_tokens_seen": 5096240, + "step": 19015 + }, + { + "epoch": 4.985583224115334, + "grad_norm": 1.2575116157531738, + "learning_rate": 2.946729837200045e-05, + "loss": 0.2486, + "num_input_tokens_seen": 5098176, + "step": 19020 + }, + { + "epoch": 4.986893840104849, + "grad_norm": 1.5065441131591797, + "learning_rate": 2.945604470499332e-05, + "loss": 0.3634, + "num_input_tokens_seen": 5099520, + "step": 19025 + }, + { + "epoch": 4.988204456094365, + "grad_norm": 4.066892623901367, + "learning_rate": 2.9444790105343735e-05, + "loss": 0.3849, + "num_input_tokens_seen": 5100688, + "step": 19030 + }, + { + "epoch": 4.989515072083879, + "grad_norm": 1.5618659257888794, + "learning_rate": 2.9433534575407257e-05, + "loss": 0.3329, + "num_input_tokens_seen": 5101888, + "step": 19035 + }, + { + "epoch": 4.990825688073395, + "grad_norm": 1.2678724527359009, + "learning_rate": 2.942227811753965e-05, + "loss": 0.3658, + "num_input_tokens_seen": 5103264, + "step": 19040 + }, + { + "epoch": 4.992136304062909, + "grad_norm": 0.7185565233230591, + "learning_rate": 2.9411020734096874e-05, + "loss": 0.3754, + "num_input_tokens_seen": 5104976, + "step": 19045 + }, + { + "epoch": 4.993446920052425, + "grad_norm": 0.8458977937698364, + "learning_rate": 2.9399762427435084e-05, + "loss": 0.3137, + "num_input_tokens_seen": 5106320, + "step": 19050 + }, + { + "epoch": 4.99475753604194, + "grad_norm": 0.7773102521896362, + "learning_rate": 2.938850319991062e-05, + "loss": 0.3961, + "num_input_tokens_seen": 5107744, + "step": 19055 + }, + { + "epoch": 4.996068152031455, + "grad_norm": 0.9285815358161926, + "learning_rate": 2.937724305388002e-05, + "loss": 0.2818, + "num_input_tokens_seen": 5109136, + "step": 19060 + }, + { + "epoch": 4.99737876802097, + "grad_norm": 0.8538596034049988, + "learning_rate": 2.936598199170001e-05, + "loss": 0.2861, + "num_input_tokens_seen": 5110544, + "step": 19065 + }, + { + "epoch": 4.9986893840104845, + "grad_norm": 1.7582753896713257, + "learning_rate": 2.9354720015727515e-05, + "loss": 0.4953, + "num_input_tokens_seen": 5111680, + "step": 19070 + }, + { + "epoch": 5.0, + "grad_norm": 1.2213753461837769, + "learning_rate": 2.934345712831964e-05, + "loss": 0.2838, + "num_input_tokens_seen": 5113312, + "step": 19075 + }, + { + "epoch": 5.001310615989515, + "grad_norm": 1.1677223443984985, + "learning_rate": 2.9332193331833696e-05, + "loss": 0.3007, + "num_input_tokens_seen": 5114800, + "step": 19080 + }, + { + "epoch": 5.001310615989515, + "eval_loss": 0.5386644601821899, + "eval_runtime": 16.8643, + "eval_samples_per_second": 50.284, + "eval_steps_per_second": 25.142, + "num_input_tokens_seen": 5114800, + "step": 19080 + }, + { + "epoch": 5.00262123197903, + "grad_norm": 1.3043981790542603, + "learning_rate": 2.9320928628627175e-05, + "loss": 0.2386, + "num_input_tokens_seen": 5116176, + "step": 19085 + }, + { + "epoch": 5.003931847968545, + "grad_norm": 1.0623791217803955, + "learning_rate": 2.9309663021057754e-05, + "loss": 0.2779, + "num_input_tokens_seen": 5117904, + "step": 19090 + }, + { + "epoch": 5.00524246395806, + "grad_norm": 1.5573606491088867, + "learning_rate": 2.929839651148331e-05, + "loss": 0.2941, + "num_input_tokens_seen": 5119520, + "step": 19095 + }, + { + "epoch": 5.006553079947575, + "grad_norm": 1.9520235061645508, + "learning_rate": 2.9287129102261907e-05, + "loss": 0.2932, + "num_input_tokens_seen": 5120688, + "step": 19100 + }, + { + "epoch": 5.007863695937091, + "grad_norm": 1.54063880443573, + "learning_rate": 2.9275860795751786e-05, + "loss": 0.4253, + "num_input_tokens_seen": 5121936, + "step": 19105 + }, + { + "epoch": 5.009174311926605, + "grad_norm": 2.839359760284424, + "learning_rate": 2.926459159431139e-05, + "loss": 0.2895, + "num_input_tokens_seen": 5122896, + "step": 19110 + }, + { + "epoch": 5.010484927916121, + "grad_norm": 0.7463701963424683, + "learning_rate": 2.9253321500299354e-05, + "loss": 0.4112, + "num_input_tokens_seen": 5124256, + "step": 19115 + }, + { + "epoch": 5.011795543905635, + "grad_norm": 1.8975896835327148, + "learning_rate": 2.9242050516074466e-05, + "loss": 0.3948, + "num_input_tokens_seen": 5125696, + "step": 19120 + }, + { + "epoch": 5.013106159895151, + "grad_norm": 0.8433171510696411, + "learning_rate": 2.9230778643995726e-05, + "loss": 0.36, + "num_input_tokens_seen": 5126960, + "step": 19125 + }, + { + "epoch": 5.014416775884666, + "grad_norm": 1.255775809288025, + "learning_rate": 2.9219505886422338e-05, + "loss": 0.3353, + "num_input_tokens_seen": 5128336, + "step": 19130 + }, + { + "epoch": 5.015727391874181, + "grad_norm": 1.796409010887146, + "learning_rate": 2.9208232245713657e-05, + "loss": 0.4492, + "num_input_tokens_seen": 5129984, + "step": 19135 + }, + { + "epoch": 5.017038007863696, + "grad_norm": 1.1270503997802734, + "learning_rate": 2.9196957724229234e-05, + "loss": 0.2318, + "num_input_tokens_seen": 5131120, + "step": 19140 + }, + { + "epoch": 5.018348623853211, + "grad_norm": 0.9310482144355774, + "learning_rate": 2.9185682324328815e-05, + "loss": 0.2438, + "num_input_tokens_seen": 5132352, + "step": 19145 + }, + { + "epoch": 5.019659239842726, + "grad_norm": 1.0727564096450806, + "learning_rate": 2.9174406048372316e-05, + "loss": 0.6217, + "num_input_tokens_seen": 5134048, + "step": 19150 + }, + { + "epoch": 5.0209698558322415, + "grad_norm": 1.638285517692566, + "learning_rate": 2.9163128898719845e-05, + "loss": 0.2995, + "num_input_tokens_seen": 5135600, + "step": 19155 + }, + { + "epoch": 5.022280471821756, + "grad_norm": 1.401483416557312, + "learning_rate": 2.915185087773169e-05, + "loss": 0.2953, + "num_input_tokens_seen": 5136784, + "step": 19160 + }, + { + "epoch": 5.023591087811272, + "grad_norm": 1.3132590055465698, + "learning_rate": 2.9140571987768316e-05, + "loss": 0.3484, + "num_input_tokens_seen": 5138656, + "step": 19165 + }, + { + "epoch": 5.024901703800786, + "grad_norm": 1.1730103492736816, + "learning_rate": 2.9129292231190386e-05, + "loss": 0.2587, + "num_input_tokens_seen": 5140256, + "step": 19170 + }, + { + "epoch": 5.026212319790301, + "grad_norm": 0.7459425926208496, + "learning_rate": 2.9118011610358726e-05, + "loss": 0.3163, + "num_input_tokens_seen": 5141648, + "step": 19175 + }, + { + "epoch": 5.027522935779817, + "grad_norm": 1.5626400709152222, + "learning_rate": 2.9106730127634346e-05, + "loss": 0.3443, + "num_input_tokens_seen": 5142848, + "step": 19180 + }, + { + "epoch": 5.028833551769331, + "grad_norm": 0.7905341982841492, + "learning_rate": 2.9095447785378443e-05, + "loss": 0.2466, + "num_input_tokens_seen": 5144128, + "step": 19185 + }, + { + "epoch": 5.030144167758847, + "grad_norm": 1.5657156705856323, + "learning_rate": 2.9084164585952405e-05, + "loss": 0.1388, + "num_input_tokens_seen": 5145808, + "step": 19190 + }, + { + "epoch": 5.031454783748361, + "grad_norm": 0.8080624341964722, + "learning_rate": 2.9072880531717773e-05, + "loss": 0.1952, + "num_input_tokens_seen": 5147728, + "step": 19195 + }, + { + "epoch": 5.032765399737877, + "grad_norm": 1.0026570558547974, + "learning_rate": 2.9061595625036277e-05, + "loss": 0.2951, + "num_input_tokens_seen": 5149200, + "step": 19200 + }, + { + "epoch": 5.034076015727392, + "grad_norm": 2.8994290828704834, + "learning_rate": 2.905030986826984e-05, + "loss": 0.3132, + "num_input_tokens_seen": 5150336, + "step": 19205 + }, + { + "epoch": 5.035386631716907, + "grad_norm": 0.7052455544471741, + "learning_rate": 2.903902326378054e-05, + "loss": 0.1629, + "num_input_tokens_seen": 5151344, + "step": 19210 + }, + { + "epoch": 5.036697247706422, + "grad_norm": 1.196686863899231, + "learning_rate": 2.9027735813930658e-05, + "loss": 0.2407, + "num_input_tokens_seen": 5152944, + "step": 19215 + }, + { + "epoch": 5.038007863695937, + "grad_norm": 1.3349459171295166, + "learning_rate": 2.9016447521082625e-05, + "loss": 0.4713, + "num_input_tokens_seen": 5154384, + "step": 19220 + }, + { + "epoch": 5.039318479685452, + "grad_norm": 1.8328489065170288, + "learning_rate": 2.9005158387599068e-05, + "loss": 0.2443, + "num_input_tokens_seen": 5155632, + "step": 19225 + }, + { + "epoch": 5.0406290956749675, + "grad_norm": 6.172170639038086, + "learning_rate": 2.8993868415842785e-05, + "loss": 0.5856, + "num_input_tokens_seen": 5156992, + "step": 19230 + }, + { + "epoch": 5.041939711664482, + "grad_norm": 2.4164488315582275, + "learning_rate": 2.8982577608176737e-05, + "loss": 0.2811, + "num_input_tokens_seen": 5158528, + "step": 19235 + }, + { + "epoch": 5.043250327653998, + "grad_norm": 1.5236918926239014, + "learning_rate": 2.8971285966964075e-05, + "loss": 0.3627, + "num_input_tokens_seen": 5160208, + "step": 19240 + }, + { + "epoch": 5.044560943643512, + "grad_norm": 0.9809311628341675, + "learning_rate": 2.8959993494568123e-05, + "loss": 0.3247, + "num_input_tokens_seen": 5161536, + "step": 19245 + }, + { + "epoch": 5.045871559633028, + "grad_norm": 1.3935139179229736, + "learning_rate": 2.894870019335238e-05, + "loss": 0.346, + "num_input_tokens_seen": 5164688, + "step": 19250 + }, + { + "epoch": 5.047182175622543, + "grad_norm": 1.2764534950256348, + "learning_rate": 2.8937406065680505e-05, + "loss": 0.3088, + "num_input_tokens_seen": 5166160, + "step": 19255 + }, + { + "epoch": 5.048492791612058, + "grad_norm": 1.394234538078308, + "learning_rate": 2.8926111113916344e-05, + "loss": 0.2715, + "num_input_tokens_seen": 5167232, + "step": 19260 + }, + { + "epoch": 5.049803407601573, + "grad_norm": 1.5566940307617188, + "learning_rate": 2.891481534042391e-05, + "loss": 0.3304, + "num_input_tokens_seen": 5168288, + "step": 19265 + }, + { + "epoch": 5.051114023591087, + "grad_norm": 1.747134804725647, + "learning_rate": 2.8903518747567394e-05, + "loss": 0.2436, + "num_input_tokens_seen": 5169344, + "step": 19270 + }, + { + "epoch": 5.052424639580603, + "grad_norm": 1.7980992794036865, + "learning_rate": 2.8892221337711143e-05, + "loss": 0.2384, + "num_input_tokens_seen": 5170288, + "step": 19275 + }, + { + "epoch": 5.053735255570118, + "grad_norm": 0.8078083395957947, + "learning_rate": 2.88809231132197e-05, + "loss": 0.2975, + "num_input_tokens_seen": 5171568, + "step": 19280 + }, + { + "epoch": 5.055045871559633, + "grad_norm": 2.4956047534942627, + "learning_rate": 2.8869624076457745e-05, + "loss": 0.2971, + "num_input_tokens_seen": 5172832, + "step": 19285 + }, + { + "epoch": 5.056356487549148, + "grad_norm": 1.296154499053955, + "learning_rate": 2.885832422979017e-05, + "loss": 0.3247, + "num_input_tokens_seen": 5174400, + "step": 19290 + }, + { + "epoch": 5.057667103538663, + "grad_norm": 1.0347633361816406, + "learning_rate": 2.8847023575581988e-05, + "loss": 0.2534, + "num_input_tokens_seen": 5175456, + "step": 19295 + }, + { + "epoch": 5.058977719528178, + "grad_norm": 1.3236372470855713, + "learning_rate": 2.8835722116198415e-05, + "loss": 0.2731, + "num_input_tokens_seen": 5176672, + "step": 19300 + }, + { + "epoch": 5.0602883355176935, + "grad_norm": 0.6885162591934204, + "learning_rate": 2.882441985400484e-05, + "loss": 0.2198, + "num_input_tokens_seen": 5177728, + "step": 19305 + }, + { + "epoch": 5.061598951507208, + "grad_norm": 0.9883347153663635, + "learning_rate": 2.88131167913668e-05, + "loss": 0.318, + "num_input_tokens_seen": 5178992, + "step": 19310 + }, + { + "epoch": 5.062909567496724, + "grad_norm": 1.0784406661987305, + "learning_rate": 2.8801812930650002e-05, + "loss": 0.21, + "num_input_tokens_seen": 5180832, + "step": 19315 + }, + { + "epoch": 5.064220183486238, + "grad_norm": 1.6023670434951782, + "learning_rate": 2.8790508274220323e-05, + "loss": 0.5233, + "num_input_tokens_seen": 5182176, + "step": 19320 + }, + { + "epoch": 5.065530799475754, + "grad_norm": 1.0051599740982056, + "learning_rate": 2.877920282444381e-05, + "loss": 0.3461, + "num_input_tokens_seen": 5183424, + "step": 19325 + }, + { + "epoch": 5.066841415465269, + "grad_norm": 1.6339914798736572, + "learning_rate": 2.8767896583686682e-05, + "loss": 0.5208, + "num_input_tokens_seen": 5184912, + "step": 19330 + }, + { + "epoch": 5.068152031454784, + "grad_norm": 0.9741342663764954, + "learning_rate": 2.875658955431531e-05, + "loss": 0.2582, + "num_input_tokens_seen": 5186512, + "step": 19335 + }, + { + "epoch": 5.069462647444299, + "grad_norm": 0.6931910514831543, + "learning_rate": 2.874528173869623e-05, + "loss": 0.203, + "num_input_tokens_seen": 5187984, + "step": 19340 + }, + { + "epoch": 5.070773263433814, + "grad_norm": 0.7753678560256958, + "learning_rate": 2.873397313919616e-05, + "loss": 0.2126, + "num_input_tokens_seen": 5189296, + "step": 19345 + }, + { + "epoch": 5.072083879423329, + "grad_norm": 1.2754132747650146, + "learning_rate": 2.872266375818196e-05, + "loss": 0.2555, + "num_input_tokens_seen": 5190400, + "step": 19350 + }, + { + "epoch": 5.073394495412844, + "grad_norm": 0.5912452340126038, + "learning_rate": 2.8711353598020662e-05, + "loss": 0.2271, + "num_input_tokens_seen": 5191408, + "step": 19355 + }, + { + "epoch": 5.074705111402359, + "grad_norm": 2.0615124702453613, + "learning_rate": 2.8700042661079472e-05, + "loss": 0.2076, + "num_input_tokens_seen": 5192672, + "step": 19360 + }, + { + "epoch": 5.076015727391874, + "grad_norm": 0.7736037373542786, + "learning_rate": 2.8688730949725752e-05, + "loss": 0.2966, + "num_input_tokens_seen": 5194112, + "step": 19365 + }, + { + "epoch": 5.077326343381389, + "grad_norm": 1.436058521270752, + "learning_rate": 2.867741846632701e-05, + "loss": 0.236, + "num_input_tokens_seen": 5195264, + "step": 19370 + }, + { + "epoch": 5.078636959370904, + "grad_norm": 1.1924771070480347, + "learning_rate": 2.8666105213250943e-05, + "loss": 0.2667, + "num_input_tokens_seen": 5197040, + "step": 19375 + }, + { + "epoch": 5.0799475753604195, + "grad_norm": 3.656508684158325, + "learning_rate": 2.8654791192865388e-05, + "loss": 0.2732, + "num_input_tokens_seen": 5198160, + "step": 19380 + }, + { + "epoch": 5.081258191349934, + "grad_norm": 1.9860578775405884, + "learning_rate": 2.864347640753835e-05, + "loss": 0.3355, + "num_input_tokens_seen": 5199120, + "step": 19385 + }, + { + "epoch": 5.08256880733945, + "grad_norm": 1.4115041494369507, + "learning_rate": 2.8632160859637998e-05, + "loss": 0.3042, + "num_input_tokens_seen": 5200544, + "step": 19390 + }, + { + "epoch": 5.083879423328964, + "grad_norm": 2.140073537826538, + "learning_rate": 2.8620844551532648e-05, + "loss": 0.2937, + "num_input_tokens_seen": 5201840, + "step": 19395 + }, + { + "epoch": 5.08519003931848, + "grad_norm": 0.9392831325531006, + "learning_rate": 2.8609527485590794e-05, + "loss": 0.2928, + "num_input_tokens_seen": 5203104, + "step": 19400 + }, + { + "epoch": 5.086500655307995, + "grad_norm": 1.8904401063919067, + "learning_rate": 2.8598209664181073e-05, + "loss": 0.3621, + "num_input_tokens_seen": 5204352, + "step": 19405 + }, + { + "epoch": 5.08781127129751, + "grad_norm": 1.604382038116455, + "learning_rate": 2.858689108967229e-05, + "loss": 0.216, + "num_input_tokens_seen": 5205920, + "step": 19410 + }, + { + "epoch": 5.089121887287025, + "grad_norm": 1.0641040802001953, + "learning_rate": 2.8575571764433396e-05, + "loss": 0.3812, + "num_input_tokens_seen": 5206992, + "step": 19415 + }, + { + "epoch": 5.09043250327654, + "grad_norm": 1.4647536277770996, + "learning_rate": 2.8564251690833505e-05, + "loss": 0.3056, + "num_input_tokens_seen": 5208400, + "step": 19420 + }, + { + "epoch": 5.091743119266055, + "grad_norm": 1.800666332244873, + "learning_rate": 2.8552930871241906e-05, + "loss": 0.4112, + "num_input_tokens_seen": 5209680, + "step": 19425 + }, + { + "epoch": 5.0930537352555705, + "grad_norm": 1.5148141384124756, + "learning_rate": 2.8541609308028006e-05, + "loss": 0.384, + "num_input_tokens_seen": 5210832, + "step": 19430 + }, + { + "epoch": 5.094364351245085, + "grad_norm": 1.5304404497146606, + "learning_rate": 2.8530287003561402e-05, + "loss": 0.2313, + "num_input_tokens_seen": 5212160, + "step": 19435 + }, + { + "epoch": 5.095674967234601, + "grad_norm": 0.9534611105918884, + "learning_rate": 2.8518963960211814e-05, + "loss": 0.3362, + "num_input_tokens_seen": 5213728, + "step": 19440 + }, + { + "epoch": 5.096985583224115, + "grad_norm": 0.6717324256896973, + "learning_rate": 2.850764018034917e-05, + "loss": 0.34, + "num_input_tokens_seen": 5215536, + "step": 19445 + }, + { + "epoch": 5.09829619921363, + "grad_norm": 1.1480549573898315, + "learning_rate": 2.8496315666343487e-05, + "loss": 0.2683, + "num_input_tokens_seen": 5216704, + "step": 19450 + }, + { + "epoch": 5.0996068152031455, + "grad_norm": 2.7842090129852295, + "learning_rate": 2.8484990420564972e-05, + "loss": 0.2494, + "num_input_tokens_seen": 5218016, + "step": 19455 + }, + { + "epoch": 5.10091743119266, + "grad_norm": 1.4737166166305542, + "learning_rate": 2.847366444538399e-05, + "loss": 0.24, + "num_input_tokens_seen": 5218960, + "step": 19460 + }, + { + "epoch": 5.102228047182176, + "grad_norm": 0.6140856146812439, + "learning_rate": 2.8462337743171046e-05, + "loss": 0.3195, + "num_input_tokens_seen": 5220352, + "step": 19465 + }, + { + "epoch": 5.10353866317169, + "grad_norm": 0.8699955940246582, + "learning_rate": 2.8451010316296783e-05, + "loss": 0.2324, + "num_input_tokens_seen": 5221776, + "step": 19470 + }, + { + "epoch": 5.104849279161206, + "grad_norm": 2.655170440673828, + "learning_rate": 2.8439682167132027e-05, + "loss": 0.5458, + "num_input_tokens_seen": 5222944, + "step": 19475 + }, + { + "epoch": 5.1061598951507206, + "grad_norm": 2.0177648067474365, + "learning_rate": 2.8428353298047738e-05, + "loss": 0.2506, + "num_input_tokens_seen": 5223904, + "step": 19480 + }, + { + "epoch": 5.107470511140236, + "grad_norm": 1.3987641334533691, + "learning_rate": 2.8417023711415036e-05, + "loss": 0.2336, + "num_input_tokens_seen": 5225408, + "step": 19485 + }, + { + "epoch": 5.108781127129751, + "grad_norm": 0.7759507298469543, + "learning_rate": 2.8405693409605166e-05, + "loss": 0.2678, + "num_input_tokens_seen": 5226368, + "step": 19490 + }, + { + "epoch": 5.110091743119266, + "grad_norm": 1.684324026107788, + "learning_rate": 2.839436239498955e-05, + "loss": 0.4728, + "num_input_tokens_seen": 5227696, + "step": 19495 + }, + { + "epoch": 5.111402359108781, + "grad_norm": 0.7443610429763794, + "learning_rate": 2.8383030669939752e-05, + "loss": 0.295, + "num_input_tokens_seen": 5229264, + "step": 19500 + }, + { + "epoch": 5.1127129750982965, + "grad_norm": 2.5188331604003906, + "learning_rate": 2.837169823682748e-05, + "loss": 0.2944, + "num_input_tokens_seen": 5230528, + "step": 19505 + }, + { + "epoch": 5.114023591087811, + "grad_norm": 3.5514674186706543, + "learning_rate": 2.8360365098024593e-05, + "loss": 0.3215, + "num_input_tokens_seen": 5231840, + "step": 19510 + }, + { + "epoch": 5.115334207077327, + "grad_norm": 1.830668568611145, + "learning_rate": 2.8349031255903098e-05, + "loss": 0.2466, + "num_input_tokens_seen": 5232912, + "step": 19515 + }, + { + "epoch": 5.116644823066841, + "grad_norm": 0.9995951056480408, + "learning_rate": 2.8337696712835143e-05, + "loss": 0.293, + "num_input_tokens_seen": 5234752, + "step": 19520 + }, + { + "epoch": 5.117955439056357, + "grad_norm": 1.9413774013519287, + "learning_rate": 2.8326361471193035e-05, + "loss": 0.4103, + "num_input_tokens_seen": 5235968, + "step": 19525 + }, + { + "epoch": 5.1192660550458715, + "grad_norm": 1.324789047241211, + "learning_rate": 2.831502553334922e-05, + "loss": 0.2423, + "num_input_tokens_seen": 5237168, + "step": 19530 + }, + { + "epoch": 5.120576671035387, + "grad_norm": 1.3077458143234253, + "learning_rate": 2.8303688901676288e-05, + "loss": 0.2342, + "num_input_tokens_seen": 5238304, + "step": 19535 + }, + { + "epoch": 5.121887287024902, + "grad_norm": 2.0307581424713135, + "learning_rate": 2.8292351578546982e-05, + "loss": 0.3271, + "num_input_tokens_seen": 5239840, + "step": 19540 + }, + { + "epoch": 5.123197903014416, + "grad_norm": 1.1052201986312866, + "learning_rate": 2.8281013566334182e-05, + "loss": 0.2762, + "num_input_tokens_seen": 5241088, + "step": 19545 + }, + { + "epoch": 5.124508519003932, + "grad_norm": 0.6804961562156677, + "learning_rate": 2.8269674867410907e-05, + "loss": 0.2577, + "num_input_tokens_seen": 5242208, + "step": 19550 + }, + { + "epoch": 5.1258191349934465, + "grad_norm": 1.2292231321334839, + "learning_rate": 2.8258335484150323e-05, + "loss": 0.2008, + "num_input_tokens_seen": 5243552, + "step": 19555 + }, + { + "epoch": 5.127129750982962, + "grad_norm": 0.8861387372016907, + "learning_rate": 2.8246995418925764e-05, + "loss": 0.2728, + "num_input_tokens_seen": 5244800, + "step": 19560 + }, + { + "epoch": 5.128440366972477, + "grad_norm": 1.5659301280975342, + "learning_rate": 2.823565467411067e-05, + "loss": 0.3377, + "num_input_tokens_seen": 5246400, + "step": 19565 + }, + { + "epoch": 5.129750982961992, + "grad_norm": 1.3959968090057373, + "learning_rate": 2.8224313252078637e-05, + "loss": 0.2667, + "num_input_tokens_seen": 5247600, + "step": 19570 + }, + { + "epoch": 5.131061598951507, + "grad_norm": 1.0457698106765747, + "learning_rate": 2.8212971155203422e-05, + "loss": 0.319, + "num_input_tokens_seen": 5248976, + "step": 19575 + }, + { + "epoch": 5.1323722149410225, + "grad_norm": 2.5565388202667236, + "learning_rate": 2.820162838585889e-05, + "loss": 0.3421, + "num_input_tokens_seen": 5249872, + "step": 19580 + }, + { + "epoch": 5.133682830930537, + "grad_norm": 1.4118380546569824, + "learning_rate": 2.819028494641907e-05, + "loss": 0.3786, + "num_input_tokens_seen": 5251376, + "step": 19585 + }, + { + "epoch": 5.134993446920053, + "grad_norm": 1.3439583778381348, + "learning_rate": 2.817894083925812e-05, + "loss": 0.3423, + "num_input_tokens_seen": 5252672, + "step": 19590 + }, + { + "epoch": 5.136304062909567, + "grad_norm": 0.7429472804069519, + "learning_rate": 2.816759606675035e-05, + "loss": 0.2341, + "num_input_tokens_seen": 5253984, + "step": 19595 + }, + { + "epoch": 5.137614678899083, + "grad_norm": 1.7175999879837036, + "learning_rate": 2.81562506312702e-05, + "loss": 0.3781, + "num_input_tokens_seen": 5255552, + "step": 19600 + }, + { + "epoch": 5.1389252948885975, + "grad_norm": 1.5965856313705444, + "learning_rate": 2.814490453519224e-05, + "loss": 0.2943, + "num_input_tokens_seen": 5256592, + "step": 19605 + }, + { + "epoch": 5.140235910878113, + "grad_norm": 1.9608306884765625, + "learning_rate": 2.8133557780891194e-05, + "loss": 0.3553, + "num_input_tokens_seen": 5257728, + "step": 19610 + }, + { + "epoch": 5.141546526867628, + "grad_norm": 1.4248968362808228, + "learning_rate": 2.812221037074192e-05, + "loss": 0.2036, + "num_input_tokens_seen": 5258816, + "step": 19615 + }, + { + "epoch": 5.142857142857143, + "grad_norm": 1.0701287984848022, + "learning_rate": 2.8110862307119413e-05, + "loss": 0.2006, + "num_input_tokens_seen": 5259904, + "step": 19620 + }, + { + "epoch": 5.144167758846658, + "grad_norm": 0.7549208402633667, + "learning_rate": 2.80995135923988e-05, + "loss": 0.2412, + "num_input_tokens_seen": 5261216, + "step": 19625 + }, + { + "epoch": 5.145478374836173, + "grad_norm": 1.0223207473754883, + "learning_rate": 2.8088164228955345e-05, + "loss": 0.1794, + "num_input_tokens_seen": 5262400, + "step": 19630 + }, + { + "epoch": 5.146788990825688, + "grad_norm": 1.324985384941101, + "learning_rate": 2.8076814219164456e-05, + "loss": 0.2454, + "num_input_tokens_seen": 5263280, + "step": 19635 + }, + { + "epoch": 5.148099606815203, + "grad_norm": 1.3441922664642334, + "learning_rate": 2.806546356540166e-05, + "loss": 0.4865, + "num_input_tokens_seen": 5264688, + "step": 19640 + }, + { + "epoch": 5.149410222804718, + "grad_norm": 1.3104984760284424, + "learning_rate": 2.8054112270042644e-05, + "loss": 0.3193, + "num_input_tokens_seen": 5265952, + "step": 19645 + }, + { + "epoch": 5.150720838794233, + "grad_norm": 0.8651281595230103, + "learning_rate": 2.8042760335463203e-05, + "loss": 0.2176, + "num_input_tokens_seen": 5267680, + "step": 19650 + }, + { + "epoch": 5.1520314547837485, + "grad_norm": 1.557101845741272, + "learning_rate": 2.8031407764039287e-05, + "loss": 0.3143, + "num_input_tokens_seen": 5268944, + "step": 19655 + }, + { + "epoch": 5.153342070773263, + "grad_norm": 1.470072627067566, + "learning_rate": 2.8020054558146964e-05, + "loss": 0.2549, + "num_input_tokens_seen": 5270096, + "step": 19660 + }, + { + "epoch": 5.154652686762779, + "grad_norm": 0.991435170173645, + "learning_rate": 2.8008700720162433e-05, + "loss": 0.5715, + "num_input_tokens_seen": 5271024, + "step": 19665 + }, + { + "epoch": 5.155963302752293, + "grad_norm": 1.1111094951629639, + "learning_rate": 2.7997346252462035e-05, + "loss": 0.3178, + "num_input_tokens_seen": 5271984, + "step": 19670 + }, + { + "epoch": 5.157273918741809, + "grad_norm": 1.3903511762619019, + "learning_rate": 2.7985991157422254e-05, + "loss": 0.2511, + "num_input_tokens_seen": 5273456, + "step": 19675 + }, + { + "epoch": 5.1585845347313235, + "grad_norm": 0.7288329005241394, + "learning_rate": 2.797463543741969e-05, + "loss": 0.2362, + "num_input_tokens_seen": 5274720, + "step": 19680 + }, + { + "epoch": 5.159895150720839, + "grad_norm": 1.2541967630386353, + "learning_rate": 2.7963279094831057e-05, + "loss": 0.3019, + "num_input_tokens_seen": 5275840, + "step": 19685 + }, + { + "epoch": 5.161205766710354, + "grad_norm": 1.8781235218048096, + "learning_rate": 2.7951922132033235e-05, + "loss": 0.2508, + "num_input_tokens_seen": 5277344, + "step": 19690 + }, + { + "epoch": 5.162516382699869, + "grad_norm": 1.1909810304641724, + "learning_rate": 2.794056455140321e-05, + "loss": 0.3241, + "num_input_tokens_seen": 5279568, + "step": 19695 + }, + { + "epoch": 5.163826998689384, + "grad_norm": 1.4038364887237549, + "learning_rate": 2.7929206355318104e-05, + "loss": 0.3341, + "num_input_tokens_seen": 5281120, + "step": 19700 + }, + { + "epoch": 5.165137614678899, + "grad_norm": 1.3546866178512573, + "learning_rate": 2.791784754615517e-05, + "loss": 0.468, + "num_input_tokens_seen": 5282816, + "step": 19705 + }, + { + "epoch": 5.166448230668414, + "grad_norm": 0.7513233423233032, + "learning_rate": 2.7906488126291784e-05, + "loss": 0.2541, + "num_input_tokens_seen": 5284144, + "step": 19710 + }, + { + "epoch": 5.16775884665793, + "grad_norm": 0.9439391493797302, + "learning_rate": 2.789512809810546e-05, + "loss": 0.2693, + "num_input_tokens_seen": 5285616, + "step": 19715 + }, + { + "epoch": 5.169069462647444, + "grad_norm": 1.637682557106018, + "learning_rate": 2.7883767463973832e-05, + "loss": 0.3194, + "num_input_tokens_seen": 5286720, + "step": 19720 + }, + { + "epoch": 5.17038007863696, + "grad_norm": 1.1431691646575928, + "learning_rate": 2.787240622627465e-05, + "loss": 0.4397, + "num_input_tokens_seen": 5287936, + "step": 19725 + }, + { + "epoch": 5.1716906946264745, + "grad_norm": 0.9844087362289429, + "learning_rate": 2.78610443873858e-05, + "loss": 0.3182, + "num_input_tokens_seen": 5289152, + "step": 19730 + }, + { + "epoch": 5.173001310615989, + "grad_norm": 1.026164174079895, + "learning_rate": 2.7849681949685314e-05, + "loss": 0.1879, + "num_input_tokens_seen": 5290160, + "step": 19735 + }, + { + "epoch": 5.174311926605505, + "grad_norm": 0.5507922172546387, + "learning_rate": 2.783831891555132e-05, + "loss": 0.2072, + "num_input_tokens_seen": 5291760, + "step": 19740 + }, + { + "epoch": 5.175622542595019, + "grad_norm": 13.00659465789795, + "learning_rate": 2.7826955287362077e-05, + "loss": 0.4708, + "num_input_tokens_seen": 5293024, + "step": 19745 + }, + { + "epoch": 5.176933158584535, + "grad_norm": 0.668633222579956, + "learning_rate": 2.7815591067495977e-05, + "loss": 0.1162, + "num_input_tokens_seen": 5293920, + "step": 19750 + }, + { + "epoch": 5.1782437745740495, + "grad_norm": 1.596430778503418, + "learning_rate": 2.7804226258331535e-05, + "loss": 0.3018, + "num_input_tokens_seen": 5295200, + "step": 19755 + }, + { + "epoch": 5.179554390563565, + "grad_norm": 0.963790237903595, + "learning_rate": 2.7792860862247377e-05, + "loss": 0.4139, + "num_input_tokens_seen": 5297088, + "step": 19760 + }, + { + "epoch": 5.18086500655308, + "grad_norm": 1.5890488624572754, + "learning_rate": 2.7781494881622268e-05, + "loss": 0.2922, + "num_input_tokens_seen": 5298480, + "step": 19765 + }, + { + "epoch": 5.182175622542595, + "grad_norm": 1.7348958253860474, + "learning_rate": 2.7770128318835087e-05, + "loss": 0.3084, + "num_input_tokens_seen": 5299648, + "step": 19770 + }, + { + "epoch": 5.18348623853211, + "grad_norm": 1.0391907691955566, + "learning_rate": 2.7758761176264835e-05, + "loss": 0.3377, + "num_input_tokens_seen": 5300752, + "step": 19775 + }, + { + "epoch": 5.184796854521625, + "grad_norm": 2.690969705581665, + "learning_rate": 2.7747393456290644e-05, + "loss": 0.3885, + "num_input_tokens_seen": 5302384, + "step": 19780 + }, + { + "epoch": 5.18610747051114, + "grad_norm": 2.39780330657959, + "learning_rate": 2.7736025161291734e-05, + "loss": 0.2874, + "num_input_tokens_seen": 5303584, + "step": 19785 + }, + { + "epoch": 5.187418086500656, + "grad_norm": 1.6330510377883911, + "learning_rate": 2.7724656293647494e-05, + "loss": 0.2495, + "num_input_tokens_seen": 5305072, + "step": 19790 + }, + { + "epoch": 5.18872870249017, + "grad_norm": 0.7298241257667542, + "learning_rate": 2.7713286855737403e-05, + "loss": 0.266, + "num_input_tokens_seen": 5306368, + "step": 19795 + }, + { + "epoch": 5.190039318479686, + "grad_norm": 0.9653077125549316, + "learning_rate": 2.7701916849941056e-05, + "loss": 0.2003, + "num_input_tokens_seen": 5307600, + "step": 19800 + }, + { + "epoch": 5.1913499344692005, + "grad_norm": 0.629037618637085, + "learning_rate": 2.7690546278638184e-05, + "loss": 0.3682, + "num_input_tokens_seen": 5308688, + "step": 19805 + }, + { + "epoch": 5.192660550458716, + "grad_norm": 4.513274192810059, + "learning_rate": 2.7679175144208626e-05, + "loss": 0.4728, + "num_input_tokens_seen": 5310192, + "step": 19810 + }, + { + "epoch": 5.193971166448231, + "grad_norm": 2.5311825275421143, + "learning_rate": 2.7667803449032333e-05, + "loss": 0.3182, + "num_input_tokens_seen": 5311392, + "step": 19815 + }, + { + "epoch": 5.195281782437746, + "grad_norm": 1.9567339420318604, + "learning_rate": 2.765643119548939e-05, + "loss": 0.227, + "num_input_tokens_seen": 5312848, + "step": 19820 + }, + { + "epoch": 5.196592398427261, + "grad_norm": 1.9435784816741943, + "learning_rate": 2.7645058385959987e-05, + "loss": 0.2831, + "num_input_tokens_seen": 5314528, + "step": 19825 + }, + { + "epoch": 5.1979030144167755, + "grad_norm": 0.8670063614845276, + "learning_rate": 2.7633685022824436e-05, + "loss": 0.2523, + "num_input_tokens_seen": 5316000, + "step": 19830 + }, + { + "epoch": 5.199213630406291, + "grad_norm": 1.474714756011963, + "learning_rate": 2.7622311108463167e-05, + "loss": 0.2993, + "num_input_tokens_seen": 5317328, + "step": 19835 + }, + { + "epoch": 5.200524246395806, + "grad_norm": 0.834121823310852, + "learning_rate": 2.76109366452567e-05, + "loss": 0.1938, + "num_input_tokens_seen": 5318336, + "step": 19840 + }, + { + "epoch": 5.201834862385321, + "grad_norm": 13.561413764953613, + "learning_rate": 2.7599561635585713e-05, + "loss": 0.9643, + "num_input_tokens_seen": 5319488, + "step": 19845 + }, + { + "epoch": 5.203145478374836, + "grad_norm": 0.6569425463676453, + "learning_rate": 2.7588186081830965e-05, + "loss": 0.3049, + "num_input_tokens_seen": 5320560, + "step": 19850 + }, + { + "epoch": 5.204456094364351, + "grad_norm": 0.8158190250396729, + "learning_rate": 2.7576809986373348e-05, + "loss": 0.249, + "num_input_tokens_seen": 5321920, + "step": 19855 + }, + { + "epoch": 5.205766710353866, + "grad_norm": 0.6879675388336182, + "learning_rate": 2.756543335159385e-05, + "loss": 0.3364, + "num_input_tokens_seen": 5323664, + "step": 19860 + }, + { + "epoch": 5.207077326343382, + "grad_norm": 1.4056875705718994, + "learning_rate": 2.7554056179873583e-05, + "loss": 0.2231, + "num_input_tokens_seen": 5324896, + "step": 19865 + }, + { + "epoch": 5.208387942332896, + "grad_norm": 1.5071018934249878, + "learning_rate": 2.7542678473593762e-05, + "loss": 0.4658, + "num_input_tokens_seen": 5326288, + "step": 19870 + }, + { + "epoch": 5.209698558322412, + "grad_norm": 1.1208438873291016, + "learning_rate": 2.753130023513574e-05, + "loss": 0.1587, + "num_input_tokens_seen": 5327472, + "step": 19875 + }, + { + "epoch": 5.2110091743119265, + "grad_norm": 1.3081214427947998, + "learning_rate": 2.7519921466880954e-05, + "loss": 0.2909, + "num_input_tokens_seen": 5328832, + "step": 19880 + }, + { + "epoch": 5.212319790301442, + "grad_norm": 1.5951191186904907, + "learning_rate": 2.7508542171210953e-05, + "loss": 0.2705, + "num_input_tokens_seen": 5329936, + "step": 19885 + }, + { + "epoch": 5.213630406290957, + "grad_norm": 1.0072383880615234, + "learning_rate": 2.7497162350507415e-05, + "loss": 0.3029, + "num_input_tokens_seen": 5331248, + "step": 19890 + }, + { + "epoch": 5.214941022280472, + "grad_norm": 1.9228394031524658, + "learning_rate": 2.7485782007152117e-05, + "loss": 0.2964, + "num_input_tokens_seen": 5332720, + "step": 19895 + }, + { + "epoch": 5.216251638269987, + "grad_norm": 0.599459707736969, + "learning_rate": 2.747440114352693e-05, + "loss": 0.3179, + "num_input_tokens_seen": 5333760, + "step": 19900 + }, + { + "epoch": 5.217562254259502, + "grad_norm": 1.5269743204116821, + "learning_rate": 2.7463019762013858e-05, + "loss": 0.5129, + "num_input_tokens_seen": 5335696, + "step": 19905 + }, + { + "epoch": 5.218872870249017, + "grad_norm": 1.283262848854065, + "learning_rate": 2.745163786499501e-05, + "loss": 0.278, + "num_input_tokens_seen": 5337088, + "step": 19910 + }, + { + "epoch": 5.220183486238533, + "grad_norm": 1.4386252164840698, + "learning_rate": 2.74402554548526e-05, + "loss": 0.2953, + "num_input_tokens_seen": 5338480, + "step": 19915 + }, + { + "epoch": 5.221494102228047, + "grad_norm": 0.983950138092041, + "learning_rate": 2.742887253396893e-05, + "loss": 0.3238, + "num_input_tokens_seen": 5339680, + "step": 19920 + }, + { + "epoch": 5.222804718217562, + "grad_norm": 2.2064602375030518, + "learning_rate": 2.7417489104726435e-05, + "loss": 0.3185, + "num_input_tokens_seen": 5340880, + "step": 19925 + }, + { + "epoch": 5.224115334207077, + "grad_norm": 1.8854471445083618, + "learning_rate": 2.740610516950765e-05, + "loss": 0.3681, + "num_input_tokens_seen": 5341888, + "step": 19930 + }, + { + "epoch": 5.225425950196592, + "grad_norm": 1.2217625379562378, + "learning_rate": 2.739472073069521e-05, + "loss": 0.2874, + "num_input_tokens_seen": 5343424, + "step": 19935 + }, + { + "epoch": 5.226736566186108, + "grad_norm": 0.9089111685752869, + "learning_rate": 2.738333579067186e-05, + "loss": 0.2464, + "num_input_tokens_seen": 5345024, + "step": 19940 + }, + { + "epoch": 5.228047182175622, + "grad_norm": 1.9039924144744873, + "learning_rate": 2.7371950351820445e-05, + "loss": 0.2768, + "num_input_tokens_seen": 5346144, + "step": 19945 + }, + { + "epoch": 5.229357798165138, + "grad_norm": 1.5026651620864868, + "learning_rate": 2.736056441652392e-05, + "loss": 0.1866, + "num_input_tokens_seen": 5346960, + "step": 19950 + }, + { + "epoch": 5.2306684141546524, + "grad_norm": 1.3893190622329712, + "learning_rate": 2.7349177987165342e-05, + "loss": 0.2984, + "num_input_tokens_seen": 5348496, + "step": 19955 + }, + { + "epoch": 5.231979030144168, + "grad_norm": 0.8350757956504822, + "learning_rate": 2.7337791066127872e-05, + "loss": 0.3447, + "num_input_tokens_seen": 5349632, + "step": 19960 + }, + { + "epoch": 5.233289646133683, + "grad_norm": 1.9125211238861084, + "learning_rate": 2.732640365579477e-05, + "loss": 0.2479, + "num_input_tokens_seen": 5350672, + "step": 19965 + }, + { + "epoch": 5.234600262123198, + "grad_norm": 1.785024642944336, + "learning_rate": 2.73150157585494e-05, + "loss": 0.2005, + "num_input_tokens_seen": 5351952, + "step": 19970 + }, + { + "epoch": 5.235910878112713, + "grad_norm": 1.0574703216552734, + "learning_rate": 2.7303627376775242e-05, + "loss": 0.2172, + "num_input_tokens_seen": 5353552, + "step": 19975 + }, + { + "epoch": 5.237221494102228, + "grad_norm": 2.5208351612091064, + "learning_rate": 2.7292238512855844e-05, + "loss": 0.4779, + "num_input_tokens_seen": 5354800, + "step": 19980 + }, + { + "epoch": 5.238532110091743, + "grad_norm": 2.771582841873169, + "learning_rate": 2.7280849169174887e-05, + "loss": 0.3049, + "num_input_tokens_seen": 5355696, + "step": 19985 + }, + { + "epoch": 5.239842726081259, + "grad_norm": 0.9395520687103271, + "learning_rate": 2.7269459348116143e-05, + "loss": 0.3215, + "num_input_tokens_seen": 5357232, + "step": 19990 + }, + { + "epoch": 5.241153342070773, + "grad_norm": 2.9561219215393066, + "learning_rate": 2.7258069052063477e-05, + "loss": 0.3191, + "num_input_tokens_seen": 5358528, + "step": 19995 + }, + { + "epoch": 5.242463958060289, + "grad_norm": 1.5800461769104004, + "learning_rate": 2.724667828340086e-05, + "loss": 0.2318, + "num_input_tokens_seen": 5359840, + "step": 20000 + }, + { + "epoch": 5.243774574049803, + "grad_norm": 6.6231231689453125, + "learning_rate": 2.7235287044512365e-05, + "loss": 0.3093, + "num_input_tokens_seen": 5360944, + "step": 20005 + }, + { + "epoch": 5.245085190039318, + "grad_norm": 1.7838332653045654, + "learning_rate": 2.7223895337782145e-05, + "loss": 0.3534, + "num_input_tokens_seen": 5362704, + "step": 20010 + }, + { + "epoch": 5.246395806028834, + "grad_norm": 2.982522964477539, + "learning_rate": 2.7212503165594483e-05, + "loss": 0.241, + "num_input_tokens_seen": 5363584, + "step": 20015 + }, + { + "epoch": 5.247706422018348, + "grad_norm": 2.6712653636932373, + "learning_rate": 2.7201110530333722e-05, + "loss": 0.2581, + "num_input_tokens_seen": 5364544, + "step": 20020 + }, + { + "epoch": 5.249017038007864, + "grad_norm": 1.6733474731445312, + "learning_rate": 2.7189717434384338e-05, + "loss": 0.3966, + "num_input_tokens_seen": 5366144, + "step": 20025 + }, + { + "epoch": 5.250327653997378, + "grad_norm": 3.554900646209717, + "learning_rate": 2.7178323880130878e-05, + "loss": 0.3475, + "num_input_tokens_seen": 5367408, + "step": 20030 + }, + { + "epoch": 5.251638269986894, + "grad_norm": 9.578316688537598, + "learning_rate": 2.716692986995799e-05, + "loss": 0.2577, + "num_input_tokens_seen": 5368480, + "step": 20035 + }, + { + "epoch": 5.252948885976409, + "grad_norm": 3.172933578491211, + "learning_rate": 2.7155535406250425e-05, + "loss": 0.426, + "num_input_tokens_seen": 5369744, + "step": 20040 + }, + { + "epoch": 5.254259501965924, + "grad_norm": 1.1352640390396118, + "learning_rate": 2.7144140491393018e-05, + "loss": 0.2653, + "num_input_tokens_seen": 5370736, + "step": 20045 + }, + { + "epoch": 5.255570117955439, + "grad_norm": 1.122785210609436, + "learning_rate": 2.713274512777072e-05, + "loss": 0.162, + "num_input_tokens_seen": 5371744, + "step": 20050 + }, + { + "epoch": 5.256880733944954, + "grad_norm": 1.1644552946090698, + "learning_rate": 2.7121349317768552e-05, + "loss": 0.4528, + "num_input_tokens_seen": 5373200, + "step": 20055 + }, + { + "epoch": 5.258191349934469, + "grad_norm": 2.875753879547119, + "learning_rate": 2.710995306377163e-05, + "loss": 0.2563, + "num_input_tokens_seen": 5374400, + "step": 20060 + }, + { + "epoch": 5.259501965923985, + "grad_norm": 1.7292324304580688, + "learning_rate": 2.7098556368165186e-05, + "loss": 0.4328, + "num_input_tokens_seen": 5375776, + "step": 20065 + }, + { + "epoch": 5.260812581913499, + "grad_norm": 1.482014536857605, + "learning_rate": 2.7087159233334512e-05, + "loss": 0.3599, + "num_input_tokens_seen": 5376848, + "step": 20070 + }, + { + "epoch": 5.262123197903015, + "grad_norm": 7.547618389129639, + "learning_rate": 2.707576166166502e-05, + "loss": 0.263, + "num_input_tokens_seen": 5378144, + "step": 20075 + }, + { + "epoch": 5.263433813892529, + "grad_norm": 3.285533905029297, + "learning_rate": 2.7064363655542203e-05, + "loss": 0.117, + "num_input_tokens_seen": 5379392, + "step": 20080 + }, + { + "epoch": 5.264744429882045, + "grad_norm": 0.3073933720588684, + "learning_rate": 2.7052965217351633e-05, + "loss": 0.4371, + "num_input_tokens_seen": 5380576, + "step": 20085 + }, + { + "epoch": 5.26605504587156, + "grad_norm": 1.0316091775894165, + "learning_rate": 2.7041566349479003e-05, + "loss": 0.1704, + "num_input_tokens_seen": 5381680, + "step": 20090 + }, + { + "epoch": 5.267365661861074, + "grad_norm": 1.3707367181777954, + "learning_rate": 2.7030167054310052e-05, + "loss": 0.3602, + "num_input_tokens_seen": 5382848, + "step": 20095 + }, + { + "epoch": 5.26867627785059, + "grad_norm": 0.9215216636657715, + "learning_rate": 2.701876733423064e-05, + "loss": 0.1882, + "num_input_tokens_seen": 5384080, + "step": 20100 + }, + { + "epoch": 5.269986893840104, + "grad_norm": 1.062743902206421, + "learning_rate": 2.700736719162672e-05, + "loss": 0.3774, + "num_input_tokens_seen": 5385920, + "step": 20105 + }, + { + "epoch": 5.27129750982962, + "grad_norm": 1.1181755065917969, + "learning_rate": 2.699596662888432e-05, + "loss": 0.3175, + "num_input_tokens_seen": 5387264, + "step": 20110 + }, + { + "epoch": 5.272608125819135, + "grad_norm": 0.7224017977714539, + "learning_rate": 2.698456564838955e-05, + "loss": 0.335, + "num_input_tokens_seen": 5388576, + "step": 20115 + }, + { + "epoch": 5.27391874180865, + "grad_norm": 1.3511148691177368, + "learning_rate": 2.6973164252528615e-05, + "loss": 0.2457, + "num_input_tokens_seen": 5389872, + "step": 20120 + }, + { + "epoch": 5.275229357798165, + "grad_norm": 1.413106083869934, + "learning_rate": 2.696176244368781e-05, + "loss": 0.2562, + "num_input_tokens_seen": 5391376, + "step": 20125 + }, + { + "epoch": 5.27653997378768, + "grad_norm": 0.9850440621376038, + "learning_rate": 2.6950360224253524e-05, + "loss": 0.3317, + "num_input_tokens_seen": 5392736, + "step": 20130 + }, + { + "epoch": 5.277850589777195, + "grad_norm": 1.753378987312317, + "learning_rate": 2.6938957596612203e-05, + "loss": 0.2752, + "num_input_tokens_seen": 5393888, + "step": 20135 + }, + { + "epoch": 5.2791612057667106, + "grad_norm": 1.179856300354004, + "learning_rate": 2.6927554563150414e-05, + "loss": 0.1466, + "num_input_tokens_seen": 5394848, + "step": 20140 + }, + { + "epoch": 5.280471821756225, + "grad_norm": 1.8025792837142944, + "learning_rate": 2.6916151126254785e-05, + "loss": 0.5164, + "num_input_tokens_seen": 5395872, + "step": 20145 + }, + { + "epoch": 5.281782437745741, + "grad_norm": 2.6527724266052246, + "learning_rate": 2.6904747288312036e-05, + "loss": 0.2962, + "num_input_tokens_seen": 5397056, + "step": 20150 + }, + { + "epoch": 5.283093053735255, + "grad_norm": 1.5301111936569214, + "learning_rate": 2.689334305170897e-05, + "loss": 0.2373, + "num_input_tokens_seen": 5398368, + "step": 20155 + }, + { + "epoch": 5.284403669724771, + "grad_norm": 1.3833134174346924, + "learning_rate": 2.688193841883247e-05, + "loss": 0.2837, + "num_input_tokens_seen": 5399664, + "step": 20160 + }, + { + "epoch": 5.285714285714286, + "grad_norm": 0.846359133720398, + "learning_rate": 2.6870533392069518e-05, + "loss": 0.2719, + "num_input_tokens_seen": 5400816, + "step": 20165 + }, + { + "epoch": 5.287024901703801, + "grad_norm": 0.9564962387084961, + "learning_rate": 2.6859127973807158e-05, + "loss": 0.2896, + "num_input_tokens_seen": 5401952, + "step": 20170 + }, + { + "epoch": 5.288335517693316, + "grad_norm": 1.5035359859466553, + "learning_rate": 2.6847722166432522e-05, + "loss": 0.2961, + "num_input_tokens_seen": 5403648, + "step": 20175 + }, + { + "epoch": 5.289646133682831, + "grad_norm": 0.8147024512290955, + "learning_rate": 2.6836315972332827e-05, + "loss": 0.3123, + "num_input_tokens_seen": 5405728, + "step": 20180 + }, + { + "epoch": 5.290956749672346, + "grad_norm": 2.5548858642578125, + "learning_rate": 2.6824909393895375e-05, + "loss": 0.2578, + "num_input_tokens_seen": 5407088, + "step": 20185 + }, + { + "epoch": 5.292267365661861, + "grad_norm": 1.2825616598129272, + "learning_rate": 2.681350243350754e-05, + "loss": 0.2943, + "num_input_tokens_seen": 5408592, + "step": 20190 + }, + { + "epoch": 5.293577981651376, + "grad_norm": 1.0617903470993042, + "learning_rate": 2.680209509355678e-05, + "loss": 0.2856, + "num_input_tokens_seen": 5409808, + "step": 20195 + }, + { + "epoch": 5.294888597640891, + "grad_norm": 1.7456754446029663, + "learning_rate": 2.679068737643063e-05, + "loss": 0.302, + "num_input_tokens_seen": 5410784, + "step": 20200 + }, + { + "epoch": 5.296199213630406, + "grad_norm": 1.265008807182312, + "learning_rate": 2.6779279284516706e-05, + "loss": 0.293, + "num_input_tokens_seen": 5411984, + "step": 20205 + }, + { + "epoch": 5.297509829619921, + "grad_norm": 1.9970022439956665, + "learning_rate": 2.6767870820202707e-05, + "loss": 0.3704, + "num_input_tokens_seen": 5413232, + "step": 20210 + }, + { + "epoch": 5.2988204456094365, + "grad_norm": 1.3124971389770508, + "learning_rate": 2.675646198587639e-05, + "loss": 0.1858, + "num_input_tokens_seen": 5414400, + "step": 20215 + }, + { + "epoch": 5.300131061598951, + "grad_norm": 1.3631315231323242, + "learning_rate": 2.6745052783925616e-05, + "loss": 0.3209, + "num_input_tokens_seen": 5415872, + "step": 20220 + }, + { + "epoch": 5.301441677588467, + "grad_norm": 4.659000873565674, + "learning_rate": 2.6733643216738325e-05, + "loss": 0.3133, + "num_input_tokens_seen": 5417152, + "step": 20225 + }, + { + "epoch": 5.302752293577981, + "grad_norm": 1.035577416419983, + "learning_rate": 2.6722233286702492e-05, + "loss": 0.2642, + "num_input_tokens_seen": 5418368, + "step": 20230 + }, + { + "epoch": 5.304062909567497, + "grad_norm": 1.11154043674469, + "learning_rate": 2.671082299620621e-05, + "loss": 0.4131, + "num_input_tokens_seen": 5419728, + "step": 20235 + }, + { + "epoch": 5.305373525557012, + "grad_norm": 1.9543440341949463, + "learning_rate": 2.6699412347637626e-05, + "loss": 0.3651, + "num_input_tokens_seen": 5421408, + "step": 20240 + }, + { + "epoch": 5.306684141546527, + "grad_norm": 1.4628820419311523, + "learning_rate": 2.6688001343384977e-05, + "loss": 0.1665, + "num_input_tokens_seen": 5422768, + "step": 20245 + }, + { + "epoch": 5.307994757536042, + "grad_norm": 1.1884251832962036, + "learning_rate": 2.6676589985836563e-05, + "loss": 0.2439, + "num_input_tokens_seen": 5424592, + "step": 20250 + }, + { + "epoch": 5.309305373525557, + "grad_norm": 1.001827359199524, + "learning_rate": 2.6665178277380766e-05, + "loss": 0.201, + "num_input_tokens_seen": 5426416, + "step": 20255 + }, + { + "epoch": 5.310615989515072, + "grad_norm": 2.4963650703430176, + "learning_rate": 2.6653766220406022e-05, + "loss": 0.4559, + "num_input_tokens_seen": 5427536, + "step": 20260 + }, + { + "epoch": 5.3119266055045875, + "grad_norm": 1.2164349555969238, + "learning_rate": 2.664235381730088e-05, + "loss": 0.1696, + "num_input_tokens_seen": 5428752, + "step": 20265 + }, + { + "epoch": 5.313237221494102, + "grad_norm": 5.87829065322876, + "learning_rate": 2.663094107045391e-05, + "loss": 0.2501, + "num_input_tokens_seen": 5429808, + "step": 20270 + }, + { + "epoch": 5.314547837483618, + "grad_norm": 1.1421806812286377, + "learning_rate": 2.6619527982253794e-05, + "loss": 0.2064, + "num_input_tokens_seen": 5431264, + "step": 20275 + }, + { + "epoch": 5.315858453473132, + "grad_norm": 1.1554975509643555, + "learning_rate": 2.6608114555089275e-05, + "loss": 0.2795, + "num_input_tokens_seen": 5432672, + "step": 20280 + }, + { + "epoch": 5.317169069462647, + "grad_norm": 1.158236026763916, + "learning_rate": 2.659670079134916e-05, + "loss": 0.2091, + "num_input_tokens_seen": 5433696, + "step": 20285 + }, + { + "epoch": 5.3184796854521625, + "grad_norm": 0.7191017866134644, + "learning_rate": 2.6585286693422322e-05, + "loss": 0.462, + "num_input_tokens_seen": 5435280, + "step": 20290 + }, + { + "epoch": 5.319790301441677, + "grad_norm": 2.1421663761138916, + "learning_rate": 2.6573872263697724e-05, + "loss": 0.4322, + "num_input_tokens_seen": 5436416, + "step": 20295 + }, + { + "epoch": 5.321100917431193, + "grad_norm": 1.0325387716293335, + "learning_rate": 2.6562457504564375e-05, + "loss": 0.1766, + "num_input_tokens_seen": 5438016, + "step": 20300 + }, + { + "epoch": 5.322411533420707, + "grad_norm": 2.043877601623535, + "learning_rate": 2.655104241841137e-05, + "loss": 0.3519, + "num_input_tokens_seen": 5439392, + "step": 20305 + }, + { + "epoch": 5.323722149410223, + "grad_norm": 0.7653983235359192, + "learning_rate": 2.653962700762787e-05, + "loss": 0.3053, + "num_input_tokens_seen": 5440960, + "step": 20310 + }, + { + "epoch": 5.325032765399738, + "grad_norm": 2.1335177421569824, + "learning_rate": 2.6528211274603092e-05, + "loss": 0.3728, + "num_input_tokens_seen": 5442816, + "step": 20315 + }, + { + "epoch": 5.326343381389253, + "grad_norm": 1.3411020040512085, + "learning_rate": 2.6516795221726344e-05, + "loss": 0.2891, + "num_input_tokens_seen": 5443792, + "step": 20320 + }, + { + "epoch": 5.327653997378768, + "grad_norm": 2.044229507446289, + "learning_rate": 2.6505378851386976e-05, + "loss": 0.4974, + "num_input_tokens_seen": 5445312, + "step": 20325 + }, + { + "epoch": 5.328964613368283, + "grad_norm": 1.091610074043274, + "learning_rate": 2.6493962165974405e-05, + "loss": 0.2708, + "num_input_tokens_seen": 5446224, + "step": 20330 + }, + { + "epoch": 5.330275229357798, + "grad_norm": 1.7226089239120483, + "learning_rate": 2.648254516787814e-05, + "loss": 0.1613, + "num_input_tokens_seen": 5447344, + "step": 20335 + }, + { + "epoch": 5.3315858453473135, + "grad_norm": 1.7974873781204224, + "learning_rate": 2.6471127859487732e-05, + "loss": 0.2226, + "num_input_tokens_seen": 5448752, + "step": 20340 + }, + { + "epoch": 5.332896461336828, + "grad_norm": 2.0845117568969727, + "learning_rate": 2.6459710243192816e-05, + "loss": 0.4, + "num_input_tokens_seen": 5450000, + "step": 20345 + }, + { + "epoch": 5.334207077326344, + "grad_norm": 1.6124688386917114, + "learning_rate": 2.6448292321383057e-05, + "loss": 0.2707, + "num_input_tokens_seen": 5451104, + "step": 20350 + }, + { + "epoch": 5.335517693315858, + "grad_norm": 1.5387911796569824, + "learning_rate": 2.6436874096448216e-05, + "loss": 0.3363, + "num_input_tokens_seen": 5452336, + "step": 20355 + }, + { + "epoch": 5.336828309305374, + "grad_norm": 2.385768175125122, + "learning_rate": 2.642545557077811e-05, + "loss": 0.3602, + "num_input_tokens_seen": 5453904, + "step": 20360 + }, + { + "epoch": 5.3381389252948885, + "grad_norm": 1.9348887205123901, + "learning_rate": 2.641403674676262e-05, + "loss": 0.3871, + "num_input_tokens_seen": 5455136, + "step": 20365 + }, + { + "epoch": 5.339449541284404, + "grad_norm": 0.701224684715271, + "learning_rate": 2.6402617626791682e-05, + "loss": 0.4327, + "num_input_tokens_seen": 5456352, + "step": 20370 + }, + { + "epoch": 5.340760157273919, + "grad_norm": 1.5298577547073364, + "learning_rate": 2.6391198213255297e-05, + "loss": 0.3677, + "num_input_tokens_seen": 5457392, + "step": 20375 + }, + { + "epoch": 5.342070773263433, + "grad_norm": 2.1986913681030273, + "learning_rate": 2.637977850854353e-05, + "loss": 0.3699, + "num_input_tokens_seen": 5459056, + "step": 20380 + }, + { + "epoch": 5.343381389252949, + "grad_norm": 1.2158777713775635, + "learning_rate": 2.6368358515046503e-05, + "loss": 0.2508, + "num_input_tokens_seen": 5460080, + "step": 20385 + }, + { + "epoch": 5.344692005242464, + "grad_norm": 1.2999861240386963, + "learning_rate": 2.6356938235154404e-05, + "loss": 0.2913, + "num_input_tokens_seen": 5461536, + "step": 20390 + }, + { + "epoch": 5.346002621231979, + "grad_norm": 2.913569450378418, + "learning_rate": 2.634551767125748e-05, + "loss": 0.2527, + "num_input_tokens_seen": 5462960, + "step": 20395 + }, + { + "epoch": 5.347313237221494, + "grad_norm": 1.603763461112976, + "learning_rate": 2.6334096825746035e-05, + "loss": 0.4531, + "num_input_tokens_seen": 5464192, + "step": 20400 + }, + { + "epoch": 5.348623853211009, + "grad_norm": 1.6681642532348633, + "learning_rate": 2.6322675701010423e-05, + "loss": 0.251, + "num_input_tokens_seen": 5465344, + "step": 20405 + }, + { + "epoch": 5.349934469200524, + "grad_norm": 2.6848628520965576, + "learning_rate": 2.6311254299441072e-05, + "loss": 0.3747, + "num_input_tokens_seen": 5466240, + "step": 20410 + }, + { + "epoch": 5.3512450851900395, + "grad_norm": 1.1311836242675781, + "learning_rate": 2.629983262342846e-05, + "loss": 0.1961, + "num_input_tokens_seen": 5467360, + "step": 20415 + }, + { + "epoch": 5.352555701179554, + "grad_norm": 2.7327468395233154, + "learning_rate": 2.628841067536314e-05, + "loss": 0.2606, + "num_input_tokens_seen": 5468512, + "step": 20420 + }, + { + "epoch": 5.35386631716907, + "grad_norm": 3.085996389389038, + "learning_rate": 2.6276988457635675e-05, + "loss": 0.3126, + "num_input_tokens_seen": 5469552, + "step": 20425 + }, + { + "epoch": 5.355176933158584, + "grad_norm": 1.2992563247680664, + "learning_rate": 2.626556597263674e-05, + "loss": 0.3528, + "num_input_tokens_seen": 5470864, + "step": 20430 + }, + { + "epoch": 5.3564875491481, + "grad_norm": 1.313973307609558, + "learning_rate": 2.625414322275703e-05, + "loss": 0.3973, + "num_input_tokens_seen": 5472080, + "step": 20435 + }, + { + "epoch": 5.3577981651376145, + "grad_norm": 1.1269595623016357, + "learning_rate": 2.6242720210387317e-05, + "loss": 0.3023, + "num_input_tokens_seen": 5473488, + "step": 20440 + }, + { + "epoch": 5.35910878112713, + "grad_norm": 0.7994669675827026, + "learning_rate": 2.623129693791841e-05, + "loss": 0.2043, + "num_input_tokens_seen": 5474640, + "step": 20445 + }, + { + "epoch": 5.360419397116645, + "grad_norm": 0.7201141119003296, + "learning_rate": 2.6219873407741175e-05, + "loss": 0.2319, + "num_input_tokens_seen": 5475808, + "step": 20450 + }, + { + "epoch": 5.36173001310616, + "grad_norm": 1.8483389616012573, + "learning_rate": 2.620844962224655e-05, + "loss": 0.2659, + "num_input_tokens_seen": 5477072, + "step": 20455 + }, + { + "epoch": 5.363040629095675, + "grad_norm": 8.031664848327637, + "learning_rate": 2.6197025583825508e-05, + "loss": 0.2951, + "num_input_tokens_seen": 5478352, + "step": 20460 + }, + { + "epoch": 5.3643512450851905, + "grad_norm": 1.6325559616088867, + "learning_rate": 2.6185601294869077e-05, + "loss": 0.1654, + "num_input_tokens_seen": 5479808, + "step": 20465 + }, + { + "epoch": 5.365661861074705, + "grad_norm": 1.4576655626296997, + "learning_rate": 2.617417675776834e-05, + "loss": 0.3029, + "num_input_tokens_seen": 5481104, + "step": 20470 + }, + { + "epoch": 5.36697247706422, + "grad_norm": 1.7467999458312988, + "learning_rate": 2.6162751974914433e-05, + "loss": 0.3222, + "num_input_tokens_seen": 5482432, + "step": 20475 + }, + { + "epoch": 5.368283093053735, + "grad_norm": 1.0884281396865845, + "learning_rate": 2.6151326948698557e-05, + "loss": 0.3962, + "num_input_tokens_seen": 5483792, + "step": 20480 + }, + { + "epoch": 5.36959370904325, + "grad_norm": 1.2426692247390747, + "learning_rate": 2.6139901681511935e-05, + "loss": 0.3906, + "num_input_tokens_seen": 5486112, + "step": 20485 + }, + { + "epoch": 5.3709043250327655, + "grad_norm": 1.4375338554382324, + "learning_rate": 2.6128476175745865e-05, + "loss": 0.2132, + "num_input_tokens_seen": 5487408, + "step": 20490 + }, + { + "epoch": 5.37221494102228, + "grad_norm": 0.6049051284790039, + "learning_rate": 2.6117050433791672e-05, + "loss": 0.2789, + "num_input_tokens_seen": 5488656, + "step": 20495 + }, + { + "epoch": 5.373525557011796, + "grad_norm": 1.0971473455429077, + "learning_rate": 2.6105624458040763e-05, + "loss": 0.3393, + "num_input_tokens_seen": 5489984, + "step": 20500 + }, + { + "epoch": 5.37483617300131, + "grad_norm": 0.9754415154457092, + "learning_rate": 2.6094198250884562e-05, + "loss": 0.3245, + "num_input_tokens_seen": 5491664, + "step": 20505 + }, + { + "epoch": 5.376146788990826, + "grad_norm": 1.6412789821624756, + "learning_rate": 2.6082771814714552e-05, + "loss": 0.3721, + "num_input_tokens_seen": 5492816, + "step": 20510 + }, + { + "epoch": 5.3774574049803405, + "grad_norm": 3.851423978805542, + "learning_rate": 2.6071345151922276e-05, + "loss": 0.3241, + "num_input_tokens_seen": 5494304, + "step": 20515 + }, + { + "epoch": 5.378768020969856, + "grad_norm": 2.5934619903564453, + "learning_rate": 2.605991826489932e-05, + "loss": 0.2924, + "num_input_tokens_seen": 5495264, + "step": 20520 + }, + { + "epoch": 5.380078636959371, + "grad_norm": 1.0546118021011353, + "learning_rate": 2.604849115603729e-05, + "loss": 0.2959, + "num_input_tokens_seen": 5496736, + "step": 20525 + }, + { + "epoch": 5.381389252948886, + "grad_norm": 1.7075200080871582, + "learning_rate": 2.6037063827727874e-05, + "loss": 0.3596, + "num_input_tokens_seen": 5497840, + "step": 20530 + }, + { + "epoch": 5.382699868938401, + "grad_norm": 0.8382020592689514, + "learning_rate": 2.6025636282362792e-05, + "loss": 0.2687, + "num_input_tokens_seen": 5498992, + "step": 20535 + }, + { + "epoch": 5.3840104849279165, + "grad_norm": 1.7623393535614014, + "learning_rate": 2.6014208522333815e-05, + "loss": 0.3107, + "num_input_tokens_seen": 5500256, + "step": 20540 + }, + { + "epoch": 5.385321100917431, + "grad_norm": 1.2629008293151855, + "learning_rate": 2.6002780550032747e-05, + "loss": 0.2891, + "num_input_tokens_seen": 5501600, + "step": 20545 + }, + { + "epoch": 5.386631716906947, + "grad_norm": 1.3298919200897217, + "learning_rate": 2.5991352367851435e-05, + "loss": 0.1694, + "num_input_tokens_seen": 5503024, + "step": 20550 + }, + { + "epoch": 5.387942332896461, + "grad_norm": 2.538440704345703, + "learning_rate": 2.5979923978181787e-05, + "loss": 0.427, + "num_input_tokens_seen": 5504352, + "step": 20555 + }, + { + "epoch": 5.389252948885977, + "grad_norm": 1.3687580823898315, + "learning_rate": 2.5968495383415742e-05, + "loss": 0.246, + "num_input_tokens_seen": 5505808, + "step": 20560 + }, + { + "epoch": 5.3905635648754915, + "grad_norm": 1.0426955223083496, + "learning_rate": 2.595706658594529e-05, + "loss": 0.4701, + "num_input_tokens_seen": 5506944, + "step": 20565 + }, + { + "epoch": 5.391874180865006, + "grad_norm": 0.8619797825813293, + "learning_rate": 2.5945637588162452e-05, + "loss": 0.2782, + "num_input_tokens_seen": 5507888, + "step": 20570 + }, + { + "epoch": 5.393184796854522, + "grad_norm": 1.7821106910705566, + "learning_rate": 2.5934208392459304e-05, + "loss": 0.2405, + "num_input_tokens_seen": 5509088, + "step": 20575 + }, + { + "epoch": 5.394495412844036, + "grad_norm": 1.1217982769012451, + "learning_rate": 2.592277900122796e-05, + "loss": 0.2678, + "num_input_tokens_seen": 5510448, + "step": 20580 + }, + { + "epoch": 5.395806028833552, + "grad_norm": 1.6364582777023315, + "learning_rate": 2.5911349416860558e-05, + "loss": 0.3006, + "num_input_tokens_seen": 5511728, + "step": 20585 + }, + { + "epoch": 5.3971166448230665, + "grad_norm": 1.6749275922775269, + "learning_rate": 2.5899919641749286e-05, + "loss": 0.292, + "num_input_tokens_seen": 5512880, + "step": 20590 + }, + { + "epoch": 5.398427260812582, + "grad_norm": 2.404024362564087, + "learning_rate": 2.588848967828641e-05, + "loss": 0.3504, + "num_input_tokens_seen": 5514032, + "step": 20595 + }, + { + "epoch": 5.399737876802097, + "grad_norm": 1.5249537229537964, + "learning_rate": 2.5877059528864168e-05, + "loss": 0.2407, + "num_input_tokens_seen": 5515504, + "step": 20600 + }, + { + "epoch": 5.401048492791612, + "grad_norm": 1.7221074104309082, + "learning_rate": 2.5865629195874886e-05, + "loss": 0.2971, + "num_input_tokens_seen": 5516864, + "step": 20605 + }, + { + "epoch": 5.402359108781127, + "grad_norm": 1.292077660560608, + "learning_rate": 2.58541986817109e-05, + "loss": 0.1821, + "num_input_tokens_seen": 5518384, + "step": 20610 + }, + { + "epoch": 5.4036697247706424, + "grad_norm": 1.4038208723068237, + "learning_rate": 2.5842767988764615e-05, + "loss": 0.325, + "num_input_tokens_seen": 5519696, + "step": 20615 + }, + { + "epoch": 5.404980340760157, + "grad_norm": 2.5015883445739746, + "learning_rate": 2.5831337119428446e-05, + "loss": 0.3364, + "num_input_tokens_seen": 5520800, + "step": 20620 + }, + { + "epoch": 5.406290956749673, + "grad_norm": 1.4058390855789185, + "learning_rate": 2.5819906076094848e-05, + "loss": 0.3421, + "num_input_tokens_seen": 5521968, + "step": 20625 + }, + { + "epoch": 5.407601572739187, + "grad_norm": 1.3582872152328491, + "learning_rate": 2.5808474861156328e-05, + "loss": 0.3649, + "num_input_tokens_seen": 5523024, + "step": 20630 + }, + { + "epoch": 5.408912188728703, + "grad_norm": 1.5584673881530762, + "learning_rate": 2.5797043477005423e-05, + "loss": 0.3785, + "num_input_tokens_seen": 5524224, + "step": 20635 + }, + { + "epoch": 5.4102228047182175, + "grad_norm": 2.8685076236724854, + "learning_rate": 2.5785611926034686e-05, + "loss": 0.3701, + "num_input_tokens_seen": 5525360, + "step": 20640 + }, + { + "epoch": 5.411533420707733, + "grad_norm": 0.11961285024881363, + "learning_rate": 2.5774180210636728e-05, + "loss": 0.2321, + "num_input_tokens_seen": 5527984, + "step": 20645 + }, + { + "epoch": 5.412844036697248, + "grad_norm": 2.513354778289795, + "learning_rate": 2.5762748333204196e-05, + "loss": 0.367, + "num_input_tokens_seen": 5529744, + "step": 20650 + }, + { + "epoch": 5.414154652686763, + "grad_norm": 1.0086430311203003, + "learning_rate": 2.575131629612976e-05, + "loss": 0.2189, + "num_input_tokens_seen": 5531152, + "step": 20655 + }, + { + "epoch": 5.415465268676278, + "grad_norm": 1.419050693511963, + "learning_rate": 2.5739884101806117e-05, + "loss": 0.2665, + "num_input_tokens_seen": 5532592, + "step": 20660 + }, + { + "epoch": 5.4167758846657925, + "grad_norm": 0.7635260820388794, + "learning_rate": 2.5728451752626008e-05, + "loss": 0.2734, + "num_input_tokens_seen": 5533824, + "step": 20665 + }, + { + "epoch": 5.418086500655308, + "grad_norm": 2.0965750217437744, + "learning_rate": 2.571701925098221e-05, + "loss": 0.499, + "num_input_tokens_seen": 5535792, + "step": 20670 + }, + { + "epoch": 5.419397116644823, + "grad_norm": 0.9754219055175781, + "learning_rate": 2.5705586599267518e-05, + "loss": 0.2957, + "num_input_tokens_seen": 5537456, + "step": 20675 + }, + { + "epoch": 5.420707732634338, + "grad_norm": 0.9048537611961365, + "learning_rate": 2.569415379987477e-05, + "loss": 0.288, + "num_input_tokens_seen": 5538688, + "step": 20680 + }, + { + "epoch": 5.422018348623853, + "grad_norm": 1.258480191230774, + "learning_rate": 2.5682720855196836e-05, + "loss": 0.2911, + "num_input_tokens_seen": 5540464, + "step": 20685 + }, + { + "epoch": 5.423328964613368, + "grad_norm": 1.4169840812683105, + "learning_rate": 2.56712877676266e-05, + "loss": 0.3399, + "num_input_tokens_seen": 5541440, + "step": 20690 + }, + { + "epoch": 5.424639580602883, + "grad_norm": 1.3226245641708374, + "learning_rate": 2.5659854539556998e-05, + "loss": 0.3346, + "num_input_tokens_seen": 5542640, + "step": 20695 + }, + { + "epoch": 5.425950196592399, + "grad_norm": 0.8139718770980835, + "learning_rate": 2.5648421173380977e-05, + "loss": 0.3151, + "num_input_tokens_seen": 5543920, + "step": 20700 + }, + { + "epoch": 5.427260812581913, + "grad_norm": 1.9094507694244385, + "learning_rate": 2.5636987671491525e-05, + "loss": 0.2786, + "num_input_tokens_seen": 5545152, + "step": 20705 + }, + { + "epoch": 5.428571428571429, + "grad_norm": 0.8769201636314392, + "learning_rate": 2.562555403628166e-05, + "loss": 0.291, + "num_input_tokens_seen": 5546336, + "step": 20710 + }, + { + "epoch": 5.4298820445609435, + "grad_norm": 0.8866831064224243, + "learning_rate": 2.5614120270144415e-05, + "loss": 0.2214, + "num_input_tokens_seen": 5547824, + "step": 20715 + }, + { + "epoch": 5.431192660550459, + "grad_norm": 3.6543362140655518, + "learning_rate": 2.5602686375472856e-05, + "loss": 0.4341, + "num_input_tokens_seen": 5548976, + "step": 20720 + }, + { + "epoch": 5.432503276539974, + "grad_norm": 2.702787399291992, + "learning_rate": 2.5591252354660083e-05, + "loss": 0.5787, + "num_input_tokens_seen": 5550160, + "step": 20725 + }, + { + "epoch": 5.433813892529489, + "grad_norm": 1.2017165422439575, + "learning_rate": 2.557981821009921e-05, + "loss": 0.3372, + "num_input_tokens_seen": 5551280, + "step": 20730 + }, + { + "epoch": 5.435124508519004, + "grad_norm": 0.8216405510902405, + "learning_rate": 2.556838394418339e-05, + "loss": 0.2054, + "num_input_tokens_seen": 5552480, + "step": 20735 + }, + { + "epoch": 5.436435124508519, + "grad_norm": 1.3203669786453247, + "learning_rate": 2.5556949559305786e-05, + "loss": 0.1989, + "num_input_tokens_seen": 5553840, + "step": 20740 + }, + { + "epoch": 5.437745740498034, + "grad_norm": 0.9096511602401733, + "learning_rate": 2.554551505785961e-05, + "loss": 0.2127, + "num_input_tokens_seen": 5555248, + "step": 20745 + }, + { + "epoch": 5.43905635648755, + "grad_norm": 0.9592581391334534, + "learning_rate": 2.553408044223807e-05, + "loss": 0.2971, + "num_input_tokens_seen": 5556592, + "step": 20750 + }, + { + "epoch": 5.440366972477064, + "grad_norm": 1.1963435411453247, + "learning_rate": 2.5522645714834426e-05, + "loss": 0.3289, + "num_input_tokens_seen": 5557888, + "step": 20755 + }, + { + "epoch": 5.441677588466579, + "grad_norm": 3.0280139446258545, + "learning_rate": 2.551121087804192e-05, + "loss": 0.2829, + "num_input_tokens_seen": 5559024, + "step": 20760 + }, + { + "epoch": 5.442988204456094, + "grad_norm": 3.926846504211426, + "learning_rate": 2.5499775934253865e-05, + "loss": 0.2284, + "num_input_tokens_seen": 5560048, + "step": 20765 + }, + { + "epoch": 5.444298820445609, + "grad_norm": 0.672614336013794, + "learning_rate": 2.5488340885863578e-05, + "loss": 0.274, + "num_input_tokens_seen": 5561296, + "step": 20770 + }, + { + "epoch": 5.445609436435125, + "grad_norm": 2.7339656352996826, + "learning_rate": 2.547690573526439e-05, + "loss": 0.352, + "num_input_tokens_seen": 5563136, + "step": 20775 + }, + { + "epoch": 5.446920052424639, + "grad_norm": 1.7068196535110474, + "learning_rate": 2.546547048484965e-05, + "loss": 0.2924, + "num_input_tokens_seen": 5564560, + "step": 20780 + }, + { + "epoch": 5.448230668414155, + "grad_norm": 1.1111469268798828, + "learning_rate": 2.545403513701274e-05, + "loss": 0.3182, + "num_input_tokens_seen": 5565696, + "step": 20785 + }, + { + "epoch": 5.4495412844036695, + "grad_norm": 2.0315372943878174, + "learning_rate": 2.544259969414706e-05, + "loss": 0.3303, + "num_input_tokens_seen": 5567296, + "step": 20790 + }, + { + "epoch": 5.450851900393185, + "grad_norm": 1.3608176708221436, + "learning_rate": 2.543116415864603e-05, + "loss": 0.3427, + "num_input_tokens_seen": 5568752, + "step": 20795 + }, + { + "epoch": 5.4521625163827, + "grad_norm": 1.923342227935791, + "learning_rate": 2.541972853290309e-05, + "loss": 0.2112, + "num_input_tokens_seen": 5569904, + "step": 20800 + }, + { + "epoch": 5.453473132372215, + "grad_norm": 1.3591806888580322, + "learning_rate": 2.540829281931169e-05, + "loss": 0.3957, + "num_input_tokens_seen": 5571568, + "step": 20805 + }, + { + "epoch": 5.45478374836173, + "grad_norm": 1.5301485061645508, + "learning_rate": 2.5396857020265307e-05, + "loss": 0.3162, + "num_input_tokens_seen": 5572704, + "step": 20810 + }, + { + "epoch": 5.456094364351245, + "grad_norm": 1.7082841396331787, + "learning_rate": 2.538542113815744e-05, + "loss": 0.3776, + "num_input_tokens_seen": 5574576, + "step": 20815 + }, + { + "epoch": 5.45740498034076, + "grad_norm": 0.38390782475471497, + "learning_rate": 2.5373985175381594e-05, + "loss": 0.2439, + "num_input_tokens_seen": 5575648, + "step": 20820 + }, + { + "epoch": 5.458715596330276, + "grad_norm": 4.916194438934326, + "learning_rate": 2.536254913433129e-05, + "loss": 0.2243, + "num_input_tokens_seen": 5576784, + "step": 20825 + }, + { + "epoch": 5.46002621231979, + "grad_norm": 1.7846169471740723, + "learning_rate": 2.535111301740009e-05, + "loss": 0.3362, + "num_input_tokens_seen": 5577920, + "step": 20830 + }, + { + "epoch": 5.461336828309306, + "grad_norm": 1.838556170463562, + "learning_rate": 2.5339676826981534e-05, + "loss": 0.4032, + "num_input_tokens_seen": 5579280, + "step": 20835 + }, + { + "epoch": 5.46264744429882, + "grad_norm": 1.1067464351654053, + "learning_rate": 2.532824056546921e-05, + "loss": 0.3308, + "num_input_tokens_seen": 5582000, + "step": 20840 + }, + { + "epoch": 5.463958060288336, + "grad_norm": 0.9999649524688721, + "learning_rate": 2.531680423525669e-05, + "loss": 0.2827, + "num_input_tokens_seen": 5583264, + "step": 20845 + }, + { + "epoch": 5.465268676277851, + "grad_norm": 2.005817413330078, + "learning_rate": 2.5305367838737608e-05, + "loss": 0.2765, + "num_input_tokens_seen": 5584304, + "step": 20850 + }, + { + "epoch": 5.466579292267365, + "grad_norm": 0.9700034856796265, + "learning_rate": 2.5293931378305558e-05, + "loss": 0.3439, + "num_input_tokens_seen": 5585824, + "step": 20855 + }, + { + "epoch": 5.467889908256881, + "grad_norm": 1.5864709615707397, + "learning_rate": 2.5282494856354176e-05, + "loss": 0.3552, + "num_input_tokens_seen": 5587152, + "step": 20860 + }, + { + "epoch": 5.4692005242463955, + "grad_norm": 1.4187238216400146, + "learning_rate": 2.5271058275277117e-05, + "loss": 0.2629, + "num_input_tokens_seen": 5588368, + "step": 20865 + }, + { + "epoch": 5.470511140235911, + "grad_norm": 1.4462419748306274, + "learning_rate": 2.5259621637468016e-05, + "loss": 0.2961, + "num_input_tokens_seen": 5590128, + "step": 20870 + }, + { + "epoch": 5.471821756225426, + "grad_norm": 1.4081816673278809, + "learning_rate": 2.5248184945320563e-05, + "loss": 0.3829, + "num_input_tokens_seen": 5591712, + "step": 20875 + }, + { + "epoch": 5.473132372214941, + "grad_norm": 2.094966173171997, + "learning_rate": 2.5236748201228428e-05, + "loss": 0.3389, + "num_input_tokens_seen": 5592880, + "step": 20880 + }, + { + "epoch": 5.474442988204456, + "grad_norm": 0.5732097029685974, + "learning_rate": 2.5225311407585306e-05, + "loss": 0.2155, + "num_input_tokens_seen": 5594064, + "step": 20885 + }, + { + "epoch": 5.475753604193971, + "grad_norm": 1.4133102893829346, + "learning_rate": 2.5213874566784902e-05, + "loss": 0.308, + "num_input_tokens_seen": 5595344, + "step": 20890 + }, + { + "epoch": 5.477064220183486, + "grad_norm": 1.4587410688400269, + "learning_rate": 2.5202437681220907e-05, + "loss": 0.3393, + "num_input_tokens_seen": 5596576, + "step": 20895 + }, + { + "epoch": 5.478374836173002, + "grad_norm": 0.9825257062911987, + "learning_rate": 2.5191000753287064e-05, + "loss": 0.2954, + "num_input_tokens_seen": 5597712, + "step": 20900 + }, + { + "epoch": 5.479685452162516, + "grad_norm": 1.143562912940979, + "learning_rate": 2.5179563785377088e-05, + "loss": 0.3336, + "num_input_tokens_seen": 5598848, + "step": 20905 + }, + { + "epoch": 5.480996068152032, + "grad_norm": 1.7063472270965576, + "learning_rate": 2.5168126779884727e-05, + "loss": 0.38, + "num_input_tokens_seen": 5600336, + "step": 20910 + }, + { + "epoch": 5.482306684141546, + "grad_norm": 0.7690276503562927, + "learning_rate": 2.5156689739203714e-05, + "loss": 0.3326, + "num_input_tokens_seen": 5601392, + "step": 20915 + }, + { + "epoch": 5.483617300131062, + "grad_norm": 1.3654255867004395, + "learning_rate": 2.5145252665727815e-05, + "loss": 0.3166, + "num_input_tokens_seen": 5602864, + "step": 20920 + }, + { + "epoch": 5.484927916120577, + "grad_norm": 1.0069034099578857, + "learning_rate": 2.5133815561850782e-05, + "loss": 0.2527, + "num_input_tokens_seen": 5604016, + "step": 20925 + }, + { + "epoch": 5.486238532110092, + "grad_norm": 1.1358870267868042, + "learning_rate": 2.5122378429966387e-05, + "loss": 0.2561, + "num_input_tokens_seen": 5605120, + "step": 20930 + }, + { + "epoch": 5.487549148099607, + "grad_norm": 2.1174612045288086, + "learning_rate": 2.511094127246839e-05, + "loss": 0.3364, + "num_input_tokens_seen": 5606256, + "step": 20935 + }, + { + "epoch": 5.488859764089122, + "grad_norm": 1.0057569742202759, + "learning_rate": 2.5099504091750585e-05, + "loss": 0.1692, + "num_input_tokens_seen": 5607424, + "step": 20940 + }, + { + "epoch": 5.490170380078637, + "grad_norm": 1.4529509544372559, + "learning_rate": 2.508806689020674e-05, + "loss": 0.491, + "num_input_tokens_seen": 5608528, + "step": 20945 + }, + { + "epoch": 5.491480996068152, + "grad_norm": 1.9734636545181274, + "learning_rate": 2.5076629670230655e-05, + "loss": 0.3308, + "num_input_tokens_seen": 5609824, + "step": 20950 + }, + { + "epoch": 5.492791612057667, + "grad_norm": 0.9185187816619873, + "learning_rate": 2.5065192434216106e-05, + "loss": 0.2936, + "num_input_tokens_seen": 5610944, + "step": 20955 + }, + { + "epoch": 5.494102228047182, + "grad_norm": 1.1689372062683105, + "learning_rate": 2.5053755184556886e-05, + "loss": 0.1303, + "num_input_tokens_seen": 5612016, + "step": 20960 + }, + { + "epoch": 5.495412844036697, + "grad_norm": 1.1462982892990112, + "learning_rate": 2.5042317923646812e-05, + "loss": 0.3344, + "num_input_tokens_seen": 5613632, + "step": 20965 + }, + { + "epoch": 5.496723460026212, + "grad_norm": 2.6418535709381104, + "learning_rate": 2.5030880653879667e-05, + "loss": 0.3584, + "num_input_tokens_seen": 5615056, + "step": 20970 + }, + { + "epoch": 5.498034076015728, + "grad_norm": 0.5416041016578674, + "learning_rate": 2.5019443377649247e-05, + "loss": 0.1738, + "num_input_tokens_seen": 5616192, + "step": 20975 + }, + { + "epoch": 5.499344692005242, + "grad_norm": 2.0033087730407715, + "learning_rate": 2.500800609734937e-05, + "loss": 0.2844, + "num_input_tokens_seen": 5617296, + "step": 20980 + }, + { + "epoch": 5.500655307994758, + "grad_norm": 1.1981490850448608, + "learning_rate": 2.4996568815373827e-05, + "loss": 0.3973, + "num_input_tokens_seen": 5618656, + "step": 20985 + }, + { + "epoch": 5.501441677588467, + "eval_loss": 0.5853312611579895, + "eval_runtime": 16.8749, + "eval_samples_per_second": 50.252, + "eval_steps_per_second": 25.126, + "num_input_tokens_seen": 5619904, + "step": 20988 + }, + { + "epoch": 5.501965923984272, + "grad_norm": 3.5010907649993896, + "learning_rate": 2.4985131534116422e-05, + "loss": 0.4481, + "num_input_tokens_seen": 5620480, + "step": 20990 + }, + { + "epoch": 5.503276539973788, + "grad_norm": 1.4268853664398193, + "learning_rate": 2.4973694255970966e-05, + "loss": 0.4728, + "num_input_tokens_seen": 5621648, + "step": 20995 + }, + { + "epoch": 5.504587155963303, + "grad_norm": 0.9307177662849426, + "learning_rate": 2.4962256983331254e-05, + "loss": 0.3087, + "num_input_tokens_seen": 5622768, + "step": 21000 + }, + { + "epoch": 5.505897771952818, + "grad_norm": 2.3600354194641113, + "learning_rate": 2.4950819718591094e-05, + "loss": 0.2689, + "num_input_tokens_seen": 5624048, + "step": 21005 + }, + { + "epoch": 5.507208387942333, + "grad_norm": 0.9889447093009949, + "learning_rate": 2.493938246414428e-05, + "loss": 0.2273, + "num_input_tokens_seen": 5625344, + "step": 21010 + }, + { + "epoch": 5.508519003931848, + "grad_norm": 1.9724712371826172, + "learning_rate": 2.4927945222384613e-05, + "loss": 0.3489, + "num_input_tokens_seen": 5626480, + "step": 21015 + }, + { + "epoch": 5.509829619921363, + "grad_norm": 1.5317888259887695, + "learning_rate": 2.491650799570588e-05, + "loss": 0.2336, + "num_input_tokens_seen": 5627744, + "step": 21020 + }, + { + "epoch": 5.511140235910878, + "grad_norm": 2.860278367996216, + "learning_rate": 2.4905070786501887e-05, + "loss": 0.2922, + "num_input_tokens_seen": 5629008, + "step": 21025 + }, + { + "epoch": 5.512450851900393, + "grad_norm": 1.0329817533493042, + "learning_rate": 2.4893633597166405e-05, + "loss": 0.3577, + "num_input_tokens_seen": 5630144, + "step": 21030 + }, + { + "epoch": 5.513761467889909, + "grad_norm": 1.6697275638580322, + "learning_rate": 2.4882196430093233e-05, + "loss": 0.3406, + "num_input_tokens_seen": 5632112, + "step": 21035 + }, + { + "epoch": 5.515072083879423, + "grad_norm": 0.8834874629974365, + "learning_rate": 2.4870759287676143e-05, + "loss": 0.322, + "num_input_tokens_seen": 5633664, + "step": 21040 + }, + { + "epoch": 5.516382699868938, + "grad_norm": 1.873725414276123, + "learning_rate": 2.485932217230892e-05, + "loss": 0.3308, + "num_input_tokens_seen": 5634992, + "step": 21045 + }, + { + "epoch": 5.517693315858454, + "grad_norm": 1.228853702545166, + "learning_rate": 2.4847885086385314e-05, + "loss": 0.2945, + "num_input_tokens_seen": 5636000, + "step": 21050 + }, + { + "epoch": 5.519003931847968, + "grad_norm": 1.6933625936508179, + "learning_rate": 2.4836448032299097e-05, + "loss": 0.3996, + "num_input_tokens_seen": 5637088, + "step": 21055 + }, + { + "epoch": 5.520314547837484, + "grad_norm": 1.0433070659637451, + "learning_rate": 2.4825011012444026e-05, + "loss": 0.3173, + "num_input_tokens_seen": 5638240, + "step": 21060 + }, + { + "epoch": 5.521625163826998, + "grad_norm": 0.800703763961792, + "learning_rate": 2.4813574029213853e-05, + "loss": 0.3687, + "num_input_tokens_seen": 5639952, + "step": 21065 + }, + { + "epoch": 5.522935779816514, + "grad_norm": 1.6005719900131226, + "learning_rate": 2.4802137085002306e-05, + "loss": 0.3523, + "num_input_tokens_seen": 5641504, + "step": 21070 + }, + { + "epoch": 5.524246395806029, + "grad_norm": 2.8372955322265625, + "learning_rate": 2.479070018220313e-05, + "loss": 0.6207, + "num_input_tokens_seen": 5642512, + "step": 21075 + }, + { + "epoch": 5.525557011795544, + "grad_norm": 0.3422592878341675, + "learning_rate": 2.4779263323210036e-05, + "loss": 0.2492, + "num_input_tokens_seen": 5643776, + "step": 21080 + }, + { + "epoch": 5.526867627785059, + "grad_norm": 1.2442295551300049, + "learning_rate": 2.4767826510416753e-05, + "loss": 0.2897, + "num_input_tokens_seen": 5645536, + "step": 21085 + }, + { + "epoch": 5.528178243774574, + "grad_norm": 1.3390158414840698, + "learning_rate": 2.4756389746216974e-05, + "loss": 0.2118, + "num_input_tokens_seen": 5646976, + "step": 21090 + }, + { + "epoch": 5.529488859764089, + "grad_norm": 1.8334423303604126, + "learning_rate": 2.4744953033004408e-05, + "loss": 0.2537, + "num_input_tokens_seen": 5648368, + "step": 21095 + }, + { + "epoch": 5.5307994757536045, + "grad_norm": 0.9113994836807251, + "learning_rate": 2.4733516373172728e-05, + "loss": 0.3028, + "num_input_tokens_seen": 5649824, + "step": 21100 + }, + { + "epoch": 5.532110091743119, + "grad_norm": 0.6390281915664673, + "learning_rate": 2.4722079769115617e-05, + "loss": 0.3143, + "num_input_tokens_seen": 5650976, + "step": 21105 + }, + { + "epoch": 5.533420707732635, + "grad_norm": 2.1058545112609863, + "learning_rate": 2.4710643223226722e-05, + "loss": 0.3593, + "num_input_tokens_seen": 5652608, + "step": 21110 + }, + { + "epoch": 5.534731323722149, + "grad_norm": 1.466262936592102, + "learning_rate": 2.4699206737899707e-05, + "loss": 0.3141, + "num_input_tokens_seen": 5653808, + "step": 21115 + }, + { + "epoch": 5.536041939711664, + "grad_norm": 0.7979567646980286, + "learning_rate": 2.4687770315528195e-05, + "loss": 0.3716, + "num_input_tokens_seen": 5655200, + "step": 21120 + }, + { + "epoch": 5.53735255570118, + "grad_norm": 1.192323923110962, + "learning_rate": 2.4676333958505827e-05, + "loss": 0.3615, + "num_input_tokens_seen": 5656192, + "step": 21125 + }, + { + "epoch": 5.538663171690695, + "grad_norm": 1.3596643209457397, + "learning_rate": 2.4664897669226193e-05, + "loss": 0.2982, + "num_input_tokens_seen": 5657472, + "step": 21130 + }, + { + "epoch": 5.53997378768021, + "grad_norm": 1.0127724409103394, + "learning_rate": 2.46534614500829e-05, + "loss": 0.3162, + "num_input_tokens_seen": 5658416, + "step": 21135 + }, + { + "epoch": 5.541284403669724, + "grad_norm": 1.3800002336502075, + "learning_rate": 2.4642025303469527e-05, + "loss": 0.2985, + "num_input_tokens_seen": 5659776, + "step": 21140 + }, + { + "epoch": 5.54259501965924, + "grad_norm": 0.9781755208969116, + "learning_rate": 2.4630589231779633e-05, + "loss": 0.2848, + "num_input_tokens_seen": 5660912, + "step": 21145 + }, + { + "epoch": 5.543905635648755, + "grad_norm": 2.5069336891174316, + "learning_rate": 2.4619153237406782e-05, + "loss": 0.338, + "num_input_tokens_seen": 5663440, + "step": 21150 + }, + { + "epoch": 5.54521625163827, + "grad_norm": 0.8163160681724548, + "learning_rate": 2.4607717322744498e-05, + "loss": 0.26, + "num_input_tokens_seen": 5664960, + "step": 21155 + }, + { + "epoch": 5.546526867627785, + "grad_norm": 0.9252485632896423, + "learning_rate": 2.4596281490186303e-05, + "loss": 0.4569, + "num_input_tokens_seen": 5666048, + "step": 21160 + }, + { + "epoch": 5.5478374836173, + "grad_norm": 1.4470824003219604, + "learning_rate": 2.4584845742125696e-05, + "loss": 0.1526, + "num_input_tokens_seen": 5667616, + "step": 21165 + }, + { + "epoch": 5.549148099606815, + "grad_norm": 1.679332971572876, + "learning_rate": 2.457341008095615e-05, + "loss": 0.2797, + "num_input_tokens_seen": 5668880, + "step": 21170 + }, + { + "epoch": 5.5504587155963305, + "grad_norm": 0.8186260461807251, + "learning_rate": 2.4561974509071143e-05, + "loss": 0.3399, + "num_input_tokens_seen": 5670032, + "step": 21175 + }, + { + "epoch": 5.551769331585845, + "grad_norm": 1.4925705194473267, + "learning_rate": 2.4550539028864113e-05, + "loss": 0.3157, + "num_input_tokens_seen": 5671280, + "step": 21180 + }, + { + "epoch": 5.553079947575361, + "grad_norm": 1.6231105327606201, + "learning_rate": 2.4539103642728494e-05, + "loss": 0.3491, + "num_input_tokens_seen": 5672544, + "step": 21185 + }, + { + "epoch": 5.554390563564875, + "grad_norm": 1.8600847721099854, + "learning_rate": 2.4527668353057683e-05, + "loss": 0.24, + "num_input_tokens_seen": 5673792, + "step": 21190 + }, + { + "epoch": 5.555701179554391, + "grad_norm": 1.3639243841171265, + "learning_rate": 2.4516233162245063e-05, + "loss": 0.4043, + "num_input_tokens_seen": 5674848, + "step": 21195 + }, + { + "epoch": 5.557011795543906, + "grad_norm": 0.8234147429466248, + "learning_rate": 2.450479807268401e-05, + "loss": 0.2824, + "num_input_tokens_seen": 5676064, + "step": 21200 + }, + { + "epoch": 5.558322411533421, + "grad_norm": 1.264941692352295, + "learning_rate": 2.449336308676786e-05, + "loss": 0.3179, + "num_input_tokens_seen": 5677472, + "step": 21205 + }, + { + "epoch": 5.559633027522936, + "grad_norm": 1.3922942876815796, + "learning_rate": 2.4481928206889946e-05, + "loss": 0.4196, + "num_input_tokens_seen": 5678752, + "step": 21210 + }, + { + "epoch": 5.56094364351245, + "grad_norm": 2.278917074203491, + "learning_rate": 2.4470493435443558e-05, + "loss": 0.458, + "num_input_tokens_seen": 5680288, + "step": 21215 + }, + { + "epoch": 5.562254259501966, + "grad_norm": 1.674106240272522, + "learning_rate": 2.4459058774821992e-05, + "loss": 0.3049, + "num_input_tokens_seen": 5681360, + "step": 21220 + }, + { + "epoch": 5.5635648754914815, + "grad_norm": 1.5248081684112549, + "learning_rate": 2.4447624227418482e-05, + "loss": 0.3133, + "num_input_tokens_seen": 5682432, + "step": 21225 + }, + { + "epoch": 5.564875491480996, + "grad_norm": 0.5878544449806213, + "learning_rate": 2.4436189795626262e-05, + "loss": 0.2163, + "num_input_tokens_seen": 5683808, + "step": 21230 + }, + { + "epoch": 5.566186107470511, + "grad_norm": 2.2229294776916504, + "learning_rate": 2.442475548183855e-05, + "loss": 0.215, + "num_input_tokens_seen": 5684928, + "step": 21235 + }, + { + "epoch": 5.567496723460026, + "grad_norm": 1.5001003742218018, + "learning_rate": 2.4413321288448518e-05, + "loss": 0.2289, + "num_input_tokens_seen": 5685840, + "step": 21240 + }, + { + "epoch": 5.568807339449541, + "grad_norm": 1.831087350845337, + "learning_rate": 2.440188721784933e-05, + "loss": 0.4088, + "num_input_tokens_seen": 5687008, + "step": 21245 + }, + { + "epoch": 5.5701179554390565, + "grad_norm": 0.8857656717300415, + "learning_rate": 2.4390453272434115e-05, + "loss": 0.2381, + "num_input_tokens_seen": 5688400, + "step": 21250 + }, + { + "epoch": 5.571428571428571, + "grad_norm": 1.65171480178833, + "learning_rate": 2.4379019454595973e-05, + "loss": 0.3395, + "num_input_tokens_seen": 5689712, + "step": 21255 + }, + { + "epoch": 5.572739187418087, + "grad_norm": 1.2011274099349976, + "learning_rate": 2.436758576672798e-05, + "loss": 0.3316, + "num_input_tokens_seen": 5690816, + "step": 21260 + }, + { + "epoch": 5.574049803407601, + "grad_norm": 0.826866865158081, + "learning_rate": 2.4356152211223198e-05, + "loss": 0.206, + "num_input_tokens_seen": 5692032, + "step": 21265 + }, + { + "epoch": 5.575360419397117, + "grad_norm": 1.2447956800460815, + "learning_rate": 2.434471879047465e-05, + "loss": 0.3851, + "num_input_tokens_seen": 5693568, + "step": 21270 + }, + { + "epoch": 5.576671035386632, + "grad_norm": 2.839433431625366, + "learning_rate": 2.4333285506875322e-05, + "loss": 0.3263, + "num_input_tokens_seen": 5695056, + "step": 21275 + }, + { + "epoch": 5.577981651376147, + "grad_norm": 4.209009647369385, + "learning_rate": 2.4321852362818195e-05, + "loss": 0.3219, + "num_input_tokens_seen": 5696480, + "step": 21280 + }, + { + "epoch": 5.579292267365662, + "grad_norm": 0.7204596996307373, + "learning_rate": 2.431041936069619e-05, + "loss": 0.2641, + "num_input_tokens_seen": 5697632, + "step": 21285 + }, + { + "epoch": 5.580602883355177, + "grad_norm": 7.896005630493164, + "learning_rate": 2.4298986502902218e-05, + "loss": 0.4728, + "num_input_tokens_seen": 5698752, + "step": 21290 + }, + { + "epoch": 5.581913499344692, + "grad_norm": 3.987064838409424, + "learning_rate": 2.4287553791829166e-05, + "loss": 0.3499, + "num_input_tokens_seen": 5700192, + "step": 21295 + }, + { + "epoch": 5.5832241153342075, + "grad_norm": 0.8918031454086304, + "learning_rate": 2.4276121229869875e-05, + "loss": 0.1937, + "num_input_tokens_seen": 5701520, + "step": 21300 + }, + { + "epoch": 5.584534731323722, + "grad_norm": 0.8146066665649414, + "learning_rate": 2.4264688819417155e-05, + "loss": 0.2655, + "num_input_tokens_seen": 5703280, + "step": 21305 + }, + { + "epoch": 5.585845347313237, + "grad_norm": 2.0594193935394287, + "learning_rate": 2.4253256562863795e-05, + "loss": 0.2037, + "num_input_tokens_seen": 5704848, + "step": 21310 + }, + { + "epoch": 5.587155963302752, + "grad_norm": 2.3188974857330322, + "learning_rate": 2.4241824462602546e-05, + "loss": 0.3529, + "num_input_tokens_seen": 5706608, + "step": 21315 + }, + { + "epoch": 5.588466579292267, + "grad_norm": 1.062041163444519, + "learning_rate": 2.4230392521026123e-05, + "loss": 0.3494, + "num_input_tokens_seen": 5707936, + "step": 21320 + }, + { + "epoch": 5.5897771952817825, + "grad_norm": 1.7193846702575684, + "learning_rate": 2.4218960740527218e-05, + "loss": 0.3191, + "num_input_tokens_seen": 5709200, + "step": 21325 + }, + { + "epoch": 5.591087811271297, + "grad_norm": 1.147909164428711, + "learning_rate": 2.4207529123498485e-05, + "loss": 0.2553, + "num_input_tokens_seen": 5710320, + "step": 21330 + }, + { + "epoch": 5.592398427260813, + "grad_norm": 0.7777418494224548, + "learning_rate": 2.4196097672332533e-05, + "loss": 0.2836, + "num_input_tokens_seen": 5711888, + "step": 21335 + }, + { + "epoch": 5.593709043250327, + "grad_norm": 1.9769526720046997, + "learning_rate": 2.4184666389421957e-05, + "loss": 0.4024, + "num_input_tokens_seen": 5713456, + "step": 21340 + }, + { + "epoch": 5.595019659239843, + "grad_norm": 1.7591789960861206, + "learning_rate": 2.417323527715929e-05, + "loss": 0.3247, + "num_input_tokens_seen": 5714720, + "step": 21345 + }, + { + "epoch": 5.5963302752293576, + "grad_norm": 1.1854335069656372, + "learning_rate": 2.416180433793705e-05, + "loss": 0.3866, + "num_input_tokens_seen": 5716048, + "step": 21350 + }, + { + "epoch": 5.597640891218873, + "grad_norm": 1.2252705097198486, + "learning_rate": 2.4150373574147715e-05, + "loss": 0.4126, + "num_input_tokens_seen": 5717840, + "step": 21355 + }, + { + "epoch": 5.598951507208388, + "grad_norm": 0.7756859660148621, + "learning_rate": 2.413894298818373e-05, + "loss": 0.2743, + "num_input_tokens_seen": 5719184, + "step": 21360 + }, + { + "epoch": 5.600262123197903, + "grad_norm": 1.2111576795578003, + "learning_rate": 2.4127512582437485e-05, + "loss": 0.2673, + "num_input_tokens_seen": 5720704, + "step": 21365 + }, + { + "epoch": 5.601572739187418, + "grad_norm": 1.7288076877593994, + "learning_rate": 2.4116082359301346e-05, + "loss": 0.1744, + "num_input_tokens_seen": 5722160, + "step": 21370 + }, + { + "epoch": 5.6028833551769335, + "grad_norm": 1.2224798202514648, + "learning_rate": 2.4104652321167638e-05, + "loss": 0.2906, + "num_input_tokens_seen": 5723392, + "step": 21375 + }, + { + "epoch": 5.604193971166448, + "grad_norm": 0.478865385055542, + "learning_rate": 2.4093222470428657e-05, + "loss": 0.2506, + "num_input_tokens_seen": 5724528, + "step": 21380 + }, + { + "epoch": 5.605504587155964, + "grad_norm": 0.5564936399459839, + "learning_rate": 2.4081792809476643e-05, + "loss": 0.2455, + "num_input_tokens_seen": 5725984, + "step": 21385 + }, + { + "epoch": 5.606815203145478, + "grad_norm": 1.84696364402771, + "learning_rate": 2.4070363340703812e-05, + "loss": 0.412, + "num_input_tokens_seen": 5727280, + "step": 21390 + }, + { + "epoch": 5.608125819134994, + "grad_norm": 1.582801103591919, + "learning_rate": 2.4058934066502333e-05, + "loss": 0.4993, + "num_input_tokens_seen": 5728336, + "step": 21395 + }, + { + "epoch": 5.6094364351245085, + "grad_norm": 2.5208165645599365, + "learning_rate": 2.4047504989264318e-05, + "loss": 0.2863, + "num_input_tokens_seen": 5729600, + "step": 21400 + }, + { + "epoch": 5.610747051114023, + "grad_norm": 2.3030595779418945, + "learning_rate": 2.4036076111381867e-05, + "loss": 0.3242, + "num_input_tokens_seen": 5731008, + "step": 21405 + }, + { + "epoch": 5.612057667103539, + "grad_norm": 1.6847739219665527, + "learning_rate": 2.4024647435247018e-05, + "loss": 0.3042, + "num_input_tokens_seen": 5732224, + "step": 21410 + }, + { + "epoch": 5.613368283093053, + "grad_norm": 1.3004217147827148, + "learning_rate": 2.4013218963251778e-05, + "loss": 0.4024, + "num_input_tokens_seen": 5733760, + "step": 21415 + }, + { + "epoch": 5.614678899082569, + "grad_norm": 0.6058666110038757, + "learning_rate": 2.4001790697788107e-05, + "loss": 0.2132, + "num_input_tokens_seen": 5735264, + "step": 21420 + }, + { + "epoch": 5.6159895150720835, + "grad_norm": 1.1000888347625732, + "learning_rate": 2.3990362641247915e-05, + "loss": 0.2326, + "num_input_tokens_seen": 5737152, + "step": 21425 + }, + { + "epoch": 5.617300131061599, + "grad_norm": 1.779268503189087, + "learning_rate": 2.3978934796023076e-05, + "loss": 0.3823, + "num_input_tokens_seen": 5738400, + "step": 21430 + }, + { + "epoch": 5.618610747051114, + "grad_norm": 0.9730377793312073, + "learning_rate": 2.3967507164505416e-05, + "loss": 0.196, + "num_input_tokens_seen": 5739632, + "step": 21435 + }, + { + "epoch": 5.619921363040629, + "grad_norm": 2.504282236099243, + "learning_rate": 2.3956079749086728e-05, + "loss": 0.3136, + "num_input_tokens_seen": 5740976, + "step": 21440 + }, + { + "epoch": 5.621231979030144, + "grad_norm": 0.6345495581626892, + "learning_rate": 2.3944652552158742e-05, + "loss": 0.298, + "num_input_tokens_seen": 5742320, + "step": 21445 + }, + { + "epoch": 5.6225425950196595, + "grad_norm": 1.687840223312378, + "learning_rate": 2.3933225576113154e-05, + "loss": 0.3339, + "num_input_tokens_seen": 5743728, + "step": 21450 + }, + { + "epoch": 5.623853211009174, + "grad_norm": 1.9448822736740112, + "learning_rate": 2.392179882334162e-05, + "loss": 0.2594, + "num_input_tokens_seen": 5744944, + "step": 21455 + }, + { + "epoch": 5.62516382699869, + "grad_norm": 1.658754587173462, + "learning_rate": 2.3910372296235717e-05, + "loss": 0.2234, + "num_input_tokens_seen": 5745856, + "step": 21460 + }, + { + "epoch": 5.626474442988204, + "grad_norm": 3.498274564743042, + "learning_rate": 2.3898945997187012e-05, + "loss": 0.3708, + "num_input_tokens_seen": 5747328, + "step": 21465 + }, + { + "epoch": 5.62778505897772, + "grad_norm": 1.6173748970031738, + "learning_rate": 2.3887519928587007e-05, + "loss": 0.5686, + "num_input_tokens_seen": 5748336, + "step": 21470 + }, + { + "epoch": 5.6290956749672345, + "grad_norm": 1.0373027324676514, + "learning_rate": 2.387609409282716e-05, + "loss": 0.302, + "num_input_tokens_seen": 5749712, + "step": 21475 + }, + { + "epoch": 5.63040629095675, + "grad_norm": 1.2246712446212769, + "learning_rate": 2.3864668492298884e-05, + "loss": 0.2689, + "num_input_tokens_seen": 5751328, + "step": 21480 + }, + { + "epoch": 5.631716906946265, + "grad_norm": 0.6422678232192993, + "learning_rate": 2.3853243129393524e-05, + "loss": 0.2482, + "num_input_tokens_seen": 5752352, + "step": 21485 + }, + { + "epoch": 5.63302752293578, + "grad_norm": 1.2762857675552368, + "learning_rate": 2.384181800650239e-05, + "loss": 0.2346, + "num_input_tokens_seen": 5753856, + "step": 21490 + }, + { + "epoch": 5.634338138925295, + "grad_norm": 0.8275046944618225, + "learning_rate": 2.383039312601675e-05, + "loss": 0.2436, + "num_input_tokens_seen": 5755216, + "step": 21495 + }, + { + "epoch": 5.6356487549148095, + "grad_norm": 1.0928537845611572, + "learning_rate": 2.3818968490327813e-05, + "loss": 0.333, + "num_input_tokens_seen": 5756512, + "step": 21500 + }, + { + "epoch": 5.636959370904325, + "grad_norm": 1.8071216344833374, + "learning_rate": 2.3807544101826722e-05, + "loss": 0.2477, + "num_input_tokens_seen": 5758160, + "step": 21505 + }, + { + "epoch": 5.63826998689384, + "grad_norm": 1.3245316743850708, + "learning_rate": 2.3796119962904597e-05, + "loss": 0.2597, + "num_input_tokens_seen": 5759776, + "step": 21510 + }, + { + "epoch": 5.639580602883355, + "grad_norm": 0.92842036485672, + "learning_rate": 2.3784696075952487e-05, + "loss": 0.1883, + "num_input_tokens_seen": 5760944, + "step": 21515 + }, + { + "epoch": 5.64089121887287, + "grad_norm": 2.9458677768707275, + "learning_rate": 2.3773272443361383e-05, + "loss": 0.3394, + "num_input_tokens_seen": 5762288, + "step": 21520 + }, + { + "epoch": 5.6422018348623855, + "grad_norm": 1.2836674451828003, + "learning_rate": 2.3761849067522235e-05, + "loss": 0.271, + "num_input_tokens_seen": 5763488, + "step": 21525 + }, + { + "epoch": 5.6435124508519, + "grad_norm": 1.714645504951477, + "learning_rate": 2.3750425950825943e-05, + "loss": 0.3849, + "num_input_tokens_seen": 5764896, + "step": 21530 + }, + { + "epoch": 5.644823066841416, + "grad_norm": 3.1618523597717285, + "learning_rate": 2.3739003095663334e-05, + "loss": 0.3264, + "num_input_tokens_seen": 5766112, + "step": 21535 + }, + { + "epoch": 5.64613368283093, + "grad_norm": 1.8465979099273682, + "learning_rate": 2.3727580504425208e-05, + "loss": 0.2407, + "num_input_tokens_seen": 5767472, + "step": 21540 + }, + { + "epoch": 5.647444298820446, + "grad_norm": 1.4730772972106934, + "learning_rate": 2.3716158179502277e-05, + "loss": 0.342, + "num_input_tokens_seen": 5769056, + "step": 21545 + }, + { + "epoch": 5.6487549148099605, + "grad_norm": 2.0571930408477783, + "learning_rate": 2.3704736123285214e-05, + "loss": 0.3598, + "num_input_tokens_seen": 5770448, + "step": 21550 + }, + { + "epoch": 5.650065530799476, + "grad_norm": 0.8703891634941101, + "learning_rate": 2.369331433816465e-05, + "loss": 0.577, + "num_input_tokens_seen": 5771808, + "step": 21555 + }, + { + "epoch": 5.651376146788991, + "grad_norm": 2.297689437866211, + "learning_rate": 2.3681892826531137e-05, + "loss": 0.2867, + "num_input_tokens_seen": 5773760, + "step": 21560 + }, + { + "epoch": 5.652686762778506, + "grad_norm": 2.111921787261963, + "learning_rate": 2.3670471590775174e-05, + "loss": 0.2359, + "num_input_tokens_seen": 5775584, + "step": 21565 + }, + { + "epoch": 5.653997378768021, + "grad_norm": 1.3818433284759521, + "learning_rate": 2.3659050633287217e-05, + "loss": 0.2432, + "num_input_tokens_seen": 5776768, + "step": 21570 + }, + { + "epoch": 5.655307994757536, + "grad_norm": 1.8946739435195923, + "learning_rate": 2.3647629956457647e-05, + "loss": 0.3746, + "num_input_tokens_seen": 5778336, + "step": 21575 + }, + { + "epoch": 5.656618610747051, + "grad_norm": 1.2704206705093384, + "learning_rate": 2.3636209562676784e-05, + "loss": 0.3095, + "num_input_tokens_seen": 5779952, + "step": 21580 + }, + { + "epoch": 5.657929226736567, + "grad_norm": 0.9956209659576416, + "learning_rate": 2.3624789454334905e-05, + "loss": 0.3847, + "num_input_tokens_seen": 5781376, + "step": 21585 + }, + { + "epoch": 5.659239842726081, + "grad_norm": 1.0318810939788818, + "learning_rate": 2.3613369633822217e-05, + "loss": 0.3014, + "num_input_tokens_seen": 5783424, + "step": 21590 + }, + { + "epoch": 5.660550458715596, + "grad_norm": 1.4432032108306885, + "learning_rate": 2.360195010352887e-05, + "loss": 0.3063, + "num_input_tokens_seen": 5784848, + "step": 21595 + }, + { + "epoch": 5.6618610747051115, + "grad_norm": 2.3602917194366455, + "learning_rate": 2.3590530865844952e-05, + "loss": 0.3197, + "num_input_tokens_seen": 5786128, + "step": 21600 + }, + { + "epoch": 5.663171690694626, + "grad_norm": 1.5272268056869507, + "learning_rate": 2.357911192316047e-05, + "loss": 0.3019, + "num_input_tokens_seen": 5787440, + "step": 21605 + }, + { + "epoch": 5.664482306684142, + "grad_norm": 0.9771504998207092, + "learning_rate": 2.356769327786543e-05, + "loss": 0.3071, + "num_input_tokens_seen": 5789056, + "step": 21610 + }, + { + "epoch": 5.665792922673656, + "grad_norm": 1.1545313596725464, + "learning_rate": 2.3556274932349703e-05, + "loss": 0.3432, + "num_input_tokens_seen": 5790416, + "step": 21615 + }, + { + "epoch": 5.667103538663172, + "grad_norm": 0.7864267230033875, + "learning_rate": 2.354485688900314e-05, + "loss": 0.3219, + "num_input_tokens_seen": 5792000, + "step": 21620 + }, + { + "epoch": 5.6684141546526865, + "grad_norm": 0.9858564734458923, + "learning_rate": 2.3533439150215517e-05, + "loss": 0.2459, + "num_input_tokens_seen": 5793792, + "step": 21625 + }, + { + "epoch": 5.669724770642202, + "grad_norm": 0.8200968503952026, + "learning_rate": 2.352202171837655e-05, + "loss": 0.2953, + "num_input_tokens_seen": 5795232, + "step": 21630 + }, + { + "epoch": 5.671035386631717, + "grad_norm": 1.0990912914276123, + "learning_rate": 2.3510604595875877e-05, + "loss": 0.2443, + "num_input_tokens_seen": 5796560, + "step": 21635 + }, + { + "epoch": 5.672346002621232, + "grad_norm": 1.3700048923492432, + "learning_rate": 2.3499187785103087e-05, + "loss": 0.2551, + "num_input_tokens_seen": 5797984, + "step": 21640 + }, + { + "epoch": 5.673656618610747, + "grad_norm": 1.2706753015518188, + "learning_rate": 2.3487771288447705e-05, + "loss": 0.2691, + "num_input_tokens_seen": 5799088, + "step": 21645 + }, + { + "epoch": 5.674967234600262, + "grad_norm": 2.036129951477051, + "learning_rate": 2.3476355108299174e-05, + "loss": 0.3829, + "num_input_tokens_seen": 5800464, + "step": 21650 + }, + { + "epoch": 5.676277850589777, + "grad_norm": 1.256600022315979, + "learning_rate": 2.3464939247046896e-05, + "loss": 0.3598, + "num_input_tokens_seen": 5802032, + "step": 21655 + }, + { + "epoch": 5.677588466579293, + "grad_norm": 0.814581036567688, + "learning_rate": 2.3453523707080165e-05, + "loss": 0.3917, + "num_input_tokens_seen": 5803424, + "step": 21660 + }, + { + "epoch": 5.678899082568807, + "grad_norm": 1.6645368337631226, + "learning_rate": 2.344210849078826e-05, + "loss": 0.2507, + "num_input_tokens_seen": 5804896, + "step": 21665 + }, + { + "epoch": 5.680209698558322, + "grad_norm": 5.029821395874023, + "learning_rate": 2.343069360056036e-05, + "loss": 0.2834, + "num_input_tokens_seen": 5806096, + "step": 21670 + }, + { + "epoch": 5.6815203145478375, + "grad_norm": 2.2405879497528076, + "learning_rate": 2.3419279038785575e-05, + "loss": 0.4172, + "num_input_tokens_seen": 5807616, + "step": 21675 + }, + { + "epoch": 5.682830930537353, + "grad_norm": 1.398659110069275, + "learning_rate": 2.3407864807852952e-05, + "loss": 0.193, + "num_input_tokens_seen": 5808768, + "step": 21680 + }, + { + "epoch": 5.684141546526868, + "grad_norm": 1.864277720451355, + "learning_rate": 2.3396450910151476e-05, + "loss": 0.2129, + "num_input_tokens_seen": 5809792, + "step": 21685 + }, + { + "epoch": 5.685452162516382, + "grad_norm": 0.8452794551849365, + "learning_rate": 2.338503734807006e-05, + "loss": 0.2838, + "num_input_tokens_seen": 5811600, + "step": 21690 + }, + { + "epoch": 5.686762778505898, + "grad_norm": 2.649940252304077, + "learning_rate": 2.337362412399753e-05, + "loss": 0.3423, + "num_input_tokens_seen": 5812736, + "step": 21695 + }, + { + "epoch": 5.6880733944954125, + "grad_norm": 1.299418568611145, + "learning_rate": 2.3362211240322663e-05, + "loss": 0.2249, + "num_input_tokens_seen": 5813600, + "step": 21700 + }, + { + "epoch": 5.689384010484928, + "grad_norm": 1.3459104299545288, + "learning_rate": 2.335079869943415e-05, + "loss": 0.3809, + "num_input_tokens_seen": 5814736, + "step": 21705 + }, + { + "epoch": 5.690694626474443, + "grad_norm": 1.370381474494934, + "learning_rate": 2.333938650372062e-05, + "loss": 0.3422, + "num_input_tokens_seen": 5816032, + "step": 21710 + }, + { + "epoch": 5.692005242463958, + "grad_norm": 1.1062253713607788, + "learning_rate": 2.3327974655570633e-05, + "loss": 0.2157, + "num_input_tokens_seen": 5816992, + "step": 21715 + }, + { + "epoch": 5.693315858453473, + "grad_norm": 2.682467460632324, + "learning_rate": 2.3316563157372645e-05, + "loss": 0.3581, + "num_input_tokens_seen": 5818032, + "step": 21720 + }, + { + "epoch": 5.694626474442988, + "grad_norm": 1.2841802835464478, + "learning_rate": 2.3305152011515098e-05, + "loss": 0.1704, + "num_input_tokens_seen": 5819232, + "step": 21725 + }, + { + "epoch": 5.695937090432503, + "grad_norm": 3.6073617935180664, + "learning_rate": 2.3293741220386296e-05, + "loss": 0.3586, + "num_input_tokens_seen": 5820608, + "step": 21730 + }, + { + "epoch": 5.697247706422019, + "grad_norm": 1.2089461088180542, + "learning_rate": 2.328233078637451e-05, + "loss": 0.2214, + "num_input_tokens_seen": 5821952, + "step": 21735 + }, + { + "epoch": 5.698558322411533, + "grad_norm": 1.0821188688278198, + "learning_rate": 2.327092071186792e-05, + "loss": 0.3962, + "num_input_tokens_seen": 5823296, + "step": 21740 + }, + { + "epoch": 5.699868938401049, + "grad_norm": 1.2461036443710327, + "learning_rate": 2.3259510999254645e-05, + "loss": 0.2468, + "num_input_tokens_seen": 5824624, + "step": 21745 + }, + { + "epoch": 5.7011795543905635, + "grad_norm": 0.7787176966667175, + "learning_rate": 2.324810165092271e-05, + "loss": 0.2429, + "num_input_tokens_seen": 5825632, + "step": 21750 + }, + { + "epoch": 5.702490170380079, + "grad_norm": 1.4002882242202759, + "learning_rate": 2.323669266926007e-05, + "loss": 0.2694, + "num_input_tokens_seen": 5827136, + "step": 21755 + }, + { + "epoch": 5.703800786369594, + "grad_norm": 1.5594573020935059, + "learning_rate": 2.3225284056654607e-05, + "loss": 0.2318, + "num_input_tokens_seen": 5828224, + "step": 21760 + }, + { + "epoch": 5.705111402359108, + "grad_norm": 0.8814515471458435, + "learning_rate": 2.3213875815494122e-05, + "loss": 0.2095, + "num_input_tokens_seen": 5829360, + "step": 21765 + }, + { + "epoch": 5.706422018348624, + "grad_norm": 0.9635298252105713, + "learning_rate": 2.3202467948166343e-05, + "loss": 0.3488, + "num_input_tokens_seen": 5830640, + "step": 21770 + }, + { + "epoch": 5.707732634338139, + "grad_norm": 0.7936224937438965, + "learning_rate": 2.319106045705891e-05, + "loss": 0.21, + "num_input_tokens_seen": 5832224, + "step": 21775 + }, + { + "epoch": 5.709043250327654, + "grad_norm": 0.8309252858161926, + "learning_rate": 2.317965334455941e-05, + "loss": 0.2263, + "num_input_tokens_seen": 5833712, + "step": 21780 + }, + { + "epoch": 5.710353866317169, + "grad_norm": 1.5102932453155518, + "learning_rate": 2.3168246613055317e-05, + "loss": 0.3879, + "num_input_tokens_seen": 5834880, + "step": 21785 + }, + { + "epoch": 5.711664482306684, + "grad_norm": 1.2937239408493042, + "learning_rate": 2.3156840264934045e-05, + "loss": 0.3042, + "num_input_tokens_seen": 5836304, + "step": 21790 + }, + { + "epoch": 5.712975098296199, + "grad_norm": 1.2566592693328857, + "learning_rate": 2.3145434302582913e-05, + "loss": 0.3668, + "num_input_tokens_seen": 5837920, + "step": 21795 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.8556939363479614, + "learning_rate": 2.313402872838918e-05, + "loss": 0.2806, + "num_input_tokens_seen": 5841344, + "step": 21800 + }, + { + "epoch": 5.715596330275229, + "grad_norm": 1.4776852130889893, + "learning_rate": 2.312262354474001e-05, + "loss": 0.4863, + "num_input_tokens_seen": 5842592, + "step": 21805 + }, + { + "epoch": 5.716906946264745, + "grad_norm": 2.096985101699829, + "learning_rate": 2.3111218754022495e-05, + "loss": 0.2469, + "num_input_tokens_seen": 5843712, + "step": 21810 + }, + { + "epoch": 5.718217562254259, + "grad_norm": 1.2997349500656128, + "learning_rate": 2.309981435862363e-05, + "loss": 0.3274, + "num_input_tokens_seen": 5845376, + "step": 21815 + }, + { + "epoch": 5.719528178243775, + "grad_norm": 2.9481000900268555, + "learning_rate": 2.308841036093033e-05, + "loss": 0.3245, + "num_input_tokens_seen": 5846896, + "step": 21820 + }, + { + "epoch": 5.7208387942332894, + "grad_norm": 2.110494613647461, + "learning_rate": 2.307700676332944e-05, + "loss": 0.4888, + "num_input_tokens_seen": 5848240, + "step": 21825 + }, + { + "epoch": 5.722149410222805, + "grad_norm": 0.7664167881011963, + "learning_rate": 2.3065603568207715e-05, + "loss": 0.2512, + "num_input_tokens_seen": 5849456, + "step": 21830 + }, + { + "epoch": 5.72346002621232, + "grad_norm": 2.6845765113830566, + "learning_rate": 2.3054200777951807e-05, + "loss": 0.2557, + "num_input_tokens_seen": 5850928, + "step": 21835 + }, + { + "epoch": 5.724770642201835, + "grad_norm": 1.4939472675323486, + "learning_rate": 2.3042798394948326e-05, + "loss": 0.2598, + "num_input_tokens_seen": 5852320, + "step": 21840 + }, + { + "epoch": 5.72608125819135, + "grad_norm": 1.0196548700332642, + "learning_rate": 2.303139642158376e-05, + "loss": 0.3535, + "num_input_tokens_seen": 5853552, + "step": 21845 + }, + { + "epoch": 5.727391874180865, + "grad_norm": 1.2190924882888794, + "learning_rate": 2.3019994860244523e-05, + "loss": 0.5215, + "num_input_tokens_seen": 5855040, + "step": 21850 + }, + { + "epoch": 5.72870249017038, + "grad_norm": 0.7399214506149292, + "learning_rate": 2.3008593713316937e-05, + "loss": 0.337, + "num_input_tokens_seen": 5857152, + "step": 21855 + }, + { + "epoch": 5.730013106159895, + "grad_norm": 2.7443060874938965, + "learning_rate": 2.2997192983187245e-05, + "loss": 0.2996, + "num_input_tokens_seen": 5858272, + "step": 21860 + }, + { + "epoch": 5.73132372214941, + "grad_norm": 1.0803338289260864, + "learning_rate": 2.2985792672241608e-05, + "loss": 0.2336, + "num_input_tokens_seen": 5859504, + "step": 21865 + }, + { + "epoch": 5.732634338138926, + "grad_norm": 1.163007378578186, + "learning_rate": 2.2974392782866078e-05, + "loss": 0.3542, + "num_input_tokens_seen": 5861072, + "step": 21870 + }, + { + "epoch": 5.73394495412844, + "grad_norm": 0.9564692974090576, + "learning_rate": 2.296299331744664e-05, + "loss": 0.2138, + "num_input_tokens_seen": 5862496, + "step": 21875 + }, + { + "epoch": 5.735255570117955, + "grad_norm": 1.2596137523651123, + "learning_rate": 2.2951594278369178e-05, + "loss": 0.353, + "num_input_tokens_seen": 5863872, + "step": 21880 + }, + { + "epoch": 5.736566186107471, + "grad_norm": 1.5313162803649902, + "learning_rate": 2.2940195668019497e-05, + "loss": 0.4898, + "num_input_tokens_seen": 5865216, + "step": 21885 + }, + { + "epoch": 5.737876802096985, + "grad_norm": 0.9011594653129578, + "learning_rate": 2.2928797488783298e-05, + "loss": 0.3085, + "num_input_tokens_seen": 5866656, + "step": 21890 + }, + { + "epoch": 5.739187418086501, + "grad_norm": 1.1754261255264282, + "learning_rate": 2.2917399743046208e-05, + "loss": 0.247, + "num_input_tokens_seen": 5867920, + "step": 21895 + }, + { + "epoch": 5.740498034076015, + "grad_norm": 1.2494269609451294, + "learning_rate": 2.2906002433193757e-05, + "loss": 0.2964, + "num_input_tokens_seen": 5869472, + "step": 21900 + }, + { + "epoch": 5.741808650065531, + "grad_norm": 0.9100179076194763, + "learning_rate": 2.2894605561611382e-05, + "loss": 0.1761, + "num_input_tokens_seen": 5870480, + "step": 21905 + }, + { + "epoch": 5.743119266055046, + "grad_norm": 1.82559072971344, + "learning_rate": 2.288320913068442e-05, + "loss": 0.4182, + "num_input_tokens_seen": 5871888, + "step": 21910 + }, + { + "epoch": 5.744429882044561, + "grad_norm": 0.8497637510299683, + "learning_rate": 2.2871813142798133e-05, + "loss": 0.2709, + "num_input_tokens_seen": 5873216, + "step": 21915 + }, + { + "epoch": 5.745740498034076, + "grad_norm": 1.7281479835510254, + "learning_rate": 2.2860417600337673e-05, + "loss": 0.2748, + "num_input_tokens_seen": 5874320, + "step": 21920 + }, + { + "epoch": 5.747051114023591, + "grad_norm": 2.0391952991485596, + "learning_rate": 2.284902250568812e-05, + "loss": 0.4158, + "num_input_tokens_seen": 5875648, + "step": 21925 + }, + { + "epoch": 5.748361730013106, + "grad_norm": 1.7978754043579102, + "learning_rate": 2.2837627861234434e-05, + "loss": 0.259, + "num_input_tokens_seen": 5876624, + "step": 21930 + }, + { + "epoch": 5.749672346002622, + "grad_norm": 1.6283518075942993, + "learning_rate": 2.2826233669361504e-05, + "loss": 0.2028, + "num_input_tokens_seen": 5877760, + "step": 21935 + }, + { + "epoch": 5.750982961992136, + "grad_norm": 1.4558968544006348, + "learning_rate": 2.2814839932454103e-05, + "loss": 0.4939, + "num_input_tokens_seen": 5879152, + "step": 21940 + }, + { + "epoch": 5.752293577981652, + "grad_norm": 3.3308842182159424, + "learning_rate": 2.2803446652896935e-05, + "loss": 0.3514, + "num_input_tokens_seen": 5880640, + "step": 21945 + }, + { + "epoch": 5.753604193971166, + "grad_norm": 2.5157980918884277, + "learning_rate": 2.279205383307458e-05, + "loss": 0.5798, + "num_input_tokens_seen": 5883504, + "step": 21950 + }, + { + "epoch": 5.754914809960681, + "grad_norm": 1.6760677099227905, + "learning_rate": 2.278066147537155e-05, + "loss": 0.3043, + "num_input_tokens_seen": 5884768, + "step": 21955 + }, + { + "epoch": 5.756225425950197, + "grad_norm": 1.6764668226242065, + "learning_rate": 2.276926958217224e-05, + "loss": 0.3436, + "num_input_tokens_seen": 5885920, + "step": 21960 + }, + { + "epoch": 5.757536041939712, + "grad_norm": 0.9270117878913879, + "learning_rate": 2.2757878155860948e-05, + "loss": 0.3424, + "num_input_tokens_seen": 5887952, + "step": 21965 + }, + { + "epoch": 5.758846657929227, + "grad_norm": 2.816688060760498, + "learning_rate": 2.2746487198821883e-05, + "loss": 0.444, + "num_input_tokens_seen": 5889872, + "step": 21970 + }, + { + "epoch": 5.760157273918741, + "grad_norm": 1.3468464612960815, + "learning_rate": 2.2735096713439156e-05, + "loss": 0.3047, + "num_input_tokens_seen": 5891408, + "step": 21975 + }, + { + "epoch": 5.761467889908257, + "grad_norm": 1.9595099687576294, + "learning_rate": 2.2723706702096774e-05, + "loss": 0.4148, + "num_input_tokens_seen": 5892448, + "step": 21980 + }, + { + "epoch": 5.762778505897772, + "grad_norm": 0.9857166409492493, + "learning_rate": 2.2712317167178648e-05, + "loss": 0.3053, + "num_input_tokens_seen": 5894208, + "step": 21985 + }, + { + "epoch": 5.764089121887287, + "grad_norm": 2.922978401184082, + "learning_rate": 2.2700928111068582e-05, + "loss": 0.2625, + "num_input_tokens_seen": 5895184, + "step": 21990 + }, + { + "epoch": 5.765399737876802, + "grad_norm": 1.0821563005447388, + "learning_rate": 2.2689539536150294e-05, + "loss": 0.3121, + "num_input_tokens_seen": 5896240, + "step": 21995 + }, + { + "epoch": 5.766710353866317, + "grad_norm": 1.4644207954406738, + "learning_rate": 2.2678151444807384e-05, + "loss": 0.4192, + "num_input_tokens_seen": 5897584, + "step": 22000 + }, + { + "epoch": 5.768020969855832, + "grad_norm": 1.113075613975525, + "learning_rate": 2.266676383942336e-05, + "loss": 0.3724, + "num_input_tokens_seen": 5898800, + "step": 22005 + }, + { + "epoch": 5.7693315858453476, + "grad_norm": 3.0247862339019775, + "learning_rate": 2.265537672238164e-05, + "loss": 0.379, + "num_input_tokens_seen": 5900080, + "step": 22010 + }, + { + "epoch": 5.770642201834862, + "grad_norm": 1.7201985120773315, + "learning_rate": 2.2643990096065518e-05, + "loss": 0.3254, + "num_input_tokens_seen": 5901200, + "step": 22015 + }, + { + "epoch": 5.771952817824378, + "grad_norm": 0.8020150065422058, + "learning_rate": 2.2632603962858207e-05, + "loss": 0.1547, + "num_input_tokens_seen": 5902416, + "step": 22020 + }, + { + "epoch": 5.773263433813892, + "grad_norm": 0.6389442086219788, + "learning_rate": 2.2621218325142788e-05, + "loss": 0.3974, + "num_input_tokens_seen": 5903712, + "step": 22025 + }, + { + "epoch": 5.774574049803408, + "grad_norm": 1.7814538478851318, + "learning_rate": 2.2609833185302262e-05, + "loss": 0.5664, + "num_input_tokens_seen": 5905216, + "step": 22030 + }, + { + "epoch": 5.775884665792923, + "grad_norm": 1.7680659294128418, + "learning_rate": 2.2598448545719525e-05, + "loss": 0.3314, + "num_input_tokens_seen": 5906576, + "step": 22035 + }, + { + "epoch": 5.777195281782438, + "grad_norm": 1.0497993230819702, + "learning_rate": 2.2587064408777352e-05, + "loss": 0.2406, + "num_input_tokens_seen": 5907648, + "step": 22040 + }, + { + "epoch": 5.778505897771953, + "grad_norm": 0.7147022485733032, + "learning_rate": 2.257568077685844e-05, + "loss": 0.43, + "num_input_tokens_seen": 5909440, + "step": 22045 + }, + { + "epoch": 5.779816513761467, + "grad_norm": 3.8128020763397217, + "learning_rate": 2.2564297652345347e-05, + "loss": 0.3771, + "num_input_tokens_seen": 5910432, + "step": 22050 + }, + { + "epoch": 5.781127129750983, + "grad_norm": 2.029412269592285, + "learning_rate": 2.255291503762054e-05, + "loss": 0.252, + "num_input_tokens_seen": 5911648, + "step": 22055 + }, + { + "epoch": 5.7824377457404985, + "grad_norm": 1.8851361274719238, + "learning_rate": 2.2541532935066394e-05, + "loss": 0.3618, + "num_input_tokens_seen": 5913168, + "step": 22060 + }, + { + "epoch": 5.783748361730013, + "grad_norm": 1.6197750568389893, + "learning_rate": 2.2530151347065153e-05, + "loss": 0.2216, + "num_input_tokens_seen": 5914528, + "step": 22065 + }, + { + "epoch": 5.785058977719528, + "grad_norm": 0.9714311957359314, + "learning_rate": 2.2518770275998967e-05, + "loss": 0.2714, + "num_input_tokens_seen": 5915968, + "step": 22070 + }, + { + "epoch": 5.786369593709043, + "grad_norm": 0.9786450266838074, + "learning_rate": 2.2507389724249878e-05, + "loss": 0.2736, + "num_input_tokens_seen": 5917184, + "step": 22075 + }, + { + "epoch": 5.787680209698558, + "grad_norm": 1.493731141090393, + "learning_rate": 2.2496009694199815e-05, + "loss": 0.3139, + "num_input_tokens_seen": 5918496, + "step": 22080 + }, + { + "epoch": 5.7889908256880735, + "grad_norm": 1.2116279602050781, + "learning_rate": 2.2484630188230596e-05, + "loss": 0.2597, + "num_input_tokens_seen": 5920096, + "step": 22085 + }, + { + "epoch": 5.790301441677588, + "grad_norm": 1.041722059249878, + "learning_rate": 2.2473251208723927e-05, + "loss": 0.4218, + "num_input_tokens_seen": 5922176, + "step": 22090 + }, + { + "epoch": 5.791612057667104, + "grad_norm": 2.0323643684387207, + "learning_rate": 2.246187275806142e-05, + "loss": 0.407, + "num_input_tokens_seen": 5923328, + "step": 22095 + }, + { + "epoch": 5.792922673656618, + "grad_norm": 0.9625224471092224, + "learning_rate": 2.2450494838624554e-05, + "loss": 0.3036, + "num_input_tokens_seen": 5924912, + "step": 22100 + }, + { + "epoch": 5.794233289646134, + "grad_norm": 1.81203031539917, + "learning_rate": 2.2439117452794723e-05, + "loss": 0.3164, + "num_input_tokens_seen": 5926512, + "step": 22105 + }, + { + "epoch": 5.795543905635649, + "grad_norm": 0.578766405582428, + "learning_rate": 2.2427740602953175e-05, + "loss": 0.2729, + "num_input_tokens_seen": 5927584, + "step": 22110 + }, + { + "epoch": 5.796854521625164, + "grad_norm": 2.148214101791382, + "learning_rate": 2.2416364291481075e-05, + "loss": 0.3773, + "num_input_tokens_seen": 5929040, + "step": 22115 + }, + { + "epoch": 5.798165137614679, + "grad_norm": 3.603466033935547, + "learning_rate": 2.240498852075946e-05, + "loss": 0.3592, + "num_input_tokens_seen": 5931664, + "step": 22120 + }, + { + "epoch": 5.799475753604194, + "grad_norm": 1.311700701713562, + "learning_rate": 2.239361329316927e-05, + "loss": 0.3059, + "num_input_tokens_seen": 5932704, + "step": 22125 + }, + { + "epoch": 5.800786369593709, + "grad_norm": 1.7584856748580933, + "learning_rate": 2.2382238611091318e-05, + "loss": 0.2489, + "num_input_tokens_seen": 5933936, + "step": 22130 + }, + { + "epoch": 5.8020969855832245, + "grad_norm": 1.171460747718811, + "learning_rate": 2.2370864476906296e-05, + "loss": 0.2747, + "num_input_tokens_seen": 5935232, + "step": 22135 + }, + { + "epoch": 5.803407601572739, + "grad_norm": 3.3608572483062744, + "learning_rate": 2.23594908929948e-05, + "loss": 0.2922, + "num_input_tokens_seen": 5936400, + "step": 22140 + }, + { + "epoch": 5.804718217562254, + "grad_norm": 2.419264078140259, + "learning_rate": 2.2348117861737297e-05, + "loss": 0.5245, + "num_input_tokens_seen": 5937648, + "step": 22145 + }, + { + "epoch": 5.806028833551769, + "grad_norm": 1.3751075267791748, + "learning_rate": 2.233674538551414e-05, + "loss": 0.2911, + "num_input_tokens_seen": 5939120, + "step": 22150 + }, + { + "epoch": 5.807339449541285, + "grad_norm": 1.101513147354126, + "learning_rate": 2.2325373466705573e-05, + "loss": 0.3958, + "num_input_tokens_seen": 5940720, + "step": 22155 + }, + { + "epoch": 5.8086500655307995, + "grad_norm": 1.0347158908843994, + "learning_rate": 2.2314002107691727e-05, + "loss": 0.3262, + "num_input_tokens_seen": 5942320, + "step": 22160 + }, + { + "epoch": 5.809960681520314, + "grad_norm": 0.9252552390098572, + "learning_rate": 2.230263131085259e-05, + "loss": 0.2438, + "num_input_tokens_seen": 5943280, + "step": 22165 + }, + { + "epoch": 5.81127129750983, + "grad_norm": 0.9744252562522888, + "learning_rate": 2.2291261078568057e-05, + "loss": 0.4411, + "num_input_tokens_seen": 5944672, + "step": 22170 + }, + { + "epoch": 5.812581913499344, + "grad_norm": 1.4277122020721436, + "learning_rate": 2.22798914132179e-05, + "loss": 0.2584, + "num_input_tokens_seen": 5946208, + "step": 22175 + }, + { + "epoch": 5.81389252948886, + "grad_norm": 1.1475859880447388, + "learning_rate": 2.226852231718177e-05, + "loss": 0.3289, + "num_input_tokens_seen": 5947456, + "step": 22180 + }, + { + "epoch": 5.815203145478375, + "grad_norm": 1.9922616481781006, + "learning_rate": 2.2257153792839202e-05, + "loss": 0.3576, + "num_input_tokens_seen": 5948560, + "step": 22185 + }, + { + "epoch": 5.81651376146789, + "grad_norm": 2.2416465282440186, + "learning_rate": 2.2245785842569604e-05, + "loss": 0.3617, + "num_input_tokens_seen": 5949504, + "step": 22190 + }, + { + "epoch": 5.817824377457405, + "grad_norm": 0.908555805683136, + "learning_rate": 2.2234418468752276e-05, + "loss": 0.4386, + "num_input_tokens_seen": 5950720, + "step": 22195 + }, + { + "epoch": 5.81913499344692, + "grad_norm": 1.5086385011672974, + "learning_rate": 2.222305167376638e-05, + "loss": 0.2512, + "num_input_tokens_seen": 5951856, + "step": 22200 + }, + { + "epoch": 5.820445609436435, + "grad_norm": 4.394117832183838, + "learning_rate": 2.221168545999097e-05, + "loss": 0.2771, + "num_input_tokens_seen": 5953184, + "step": 22205 + }, + { + "epoch": 5.8217562254259505, + "grad_norm": 1.897745966911316, + "learning_rate": 2.2200319829804978e-05, + "loss": 0.2769, + "num_input_tokens_seen": 5954208, + "step": 22210 + }, + { + "epoch": 5.823066841415465, + "grad_norm": 1.2615376710891724, + "learning_rate": 2.2188954785587207e-05, + "loss": 0.199, + "num_input_tokens_seen": 5955376, + "step": 22215 + }, + { + "epoch": 5.824377457404981, + "grad_norm": 1.0133715867996216, + "learning_rate": 2.2177590329716345e-05, + "loss": 0.2959, + "num_input_tokens_seen": 5956752, + "step": 22220 + }, + { + "epoch": 5.825688073394495, + "grad_norm": 0.8986086249351501, + "learning_rate": 2.2166226464570947e-05, + "loss": 0.1849, + "num_input_tokens_seen": 5957760, + "step": 22225 + }, + { + "epoch": 5.826998689384011, + "grad_norm": 1.2058181762695312, + "learning_rate": 2.2154863192529453e-05, + "loss": 0.3648, + "num_input_tokens_seen": 5959152, + "step": 22230 + }, + { + "epoch": 5.8283093053735255, + "grad_norm": 2.0689072608947754, + "learning_rate": 2.2143500515970173e-05, + "loss": 0.2452, + "num_input_tokens_seen": 5960368, + "step": 22235 + }, + { + "epoch": 5.82961992136304, + "grad_norm": 0.8836621046066284, + "learning_rate": 2.21321384372713e-05, + "loss": 0.3072, + "num_input_tokens_seen": 5962016, + "step": 22240 + }, + { + "epoch": 5.830930537352556, + "grad_norm": 0.04678456112742424, + "learning_rate": 2.2120776958810902e-05, + "loss": 0.1924, + "num_input_tokens_seen": 5963040, + "step": 22245 + }, + { + "epoch": 5.832241153342071, + "grad_norm": 1.0426108837127686, + "learning_rate": 2.2109416082966907e-05, + "loss": 0.2743, + "num_input_tokens_seen": 5964256, + "step": 22250 + }, + { + "epoch": 5.833551769331586, + "grad_norm": 2.774670124053955, + "learning_rate": 2.2098055812117143e-05, + "loss": 0.2723, + "num_input_tokens_seen": 5965200, + "step": 22255 + }, + { + "epoch": 5.834862385321101, + "grad_norm": 1.4949978590011597, + "learning_rate": 2.208669614863927e-05, + "loss": 0.3606, + "num_input_tokens_seen": 5966784, + "step": 22260 + }, + { + "epoch": 5.836173001310616, + "grad_norm": 2.4641830921173096, + "learning_rate": 2.2075337094910868e-05, + "loss": 0.4028, + "num_input_tokens_seen": 5967872, + "step": 22265 + }, + { + "epoch": 5.837483617300131, + "grad_norm": 2.5038702487945557, + "learning_rate": 2.2063978653309354e-05, + "loss": 0.4015, + "num_input_tokens_seen": 5969648, + "step": 22270 + }, + { + "epoch": 5.838794233289646, + "grad_norm": 1.2349507808685303, + "learning_rate": 2.2052620826212035e-05, + "loss": 0.4195, + "num_input_tokens_seen": 5971168, + "step": 22275 + }, + { + "epoch": 5.840104849279161, + "grad_norm": 1.0257847309112549, + "learning_rate": 2.204126361599609e-05, + "loss": 0.3831, + "num_input_tokens_seen": 5973200, + "step": 22280 + }, + { + "epoch": 5.8414154652686765, + "grad_norm": 2.3430416584014893, + "learning_rate": 2.2029907025038548e-05, + "loss": 0.224, + "num_input_tokens_seen": 5974224, + "step": 22285 + }, + { + "epoch": 5.842726081258191, + "grad_norm": 1.396978735923767, + "learning_rate": 2.201855105571634e-05, + "loss": 0.3928, + "num_input_tokens_seen": 5975504, + "step": 22290 + }, + { + "epoch": 5.844036697247707, + "grad_norm": 1.8615670204162598, + "learning_rate": 2.2007195710406236e-05, + "loss": 0.261, + "num_input_tokens_seen": 5976608, + "step": 22295 + }, + { + "epoch": 5.845347313237221, + "grad_norm": 2.487358331680298, + "learning_rate": 2.1995840991484906e-05, + "loss": 0.3355, + "num_input_tokens_seen": 5977760, + "step": 22300 + }, + { + "epoch": 5.846657929226737, + "grad_norm": 1.3990964889526367, + "learning_rate": 2.198448690132886e-05, + "loss": 0.2618, + "num_input_tokens_seen": 5978896, + "step": 22305 + }, + { + "epoch": 5.8479685452162515, + "grad_norm": 0.8850935101509094, + "learning_rate": 2.19731334423145e-05, + "loss": 0.3501, + "num_input_tokens_seen": 5980576, + "step": 22310 + }, + { + "epoch": 5.849279161205767, + "grad_norm": 1.2334940433502197, + "learning_rate": 2.196178061681809e-05, + "loss": 0.3858, + "num_input_tokens_seen": 5981632, + "step": 22315 + }, + { + "epoch": 5.850589777195282, + "grad_norm": 1.1159132719039917, + "learning_rate": 2.1950428427215737e-05, + "loss": 0.2392, + "num_input_tokens_seen": 5983392, + "step": 22320 + }, + { + "epoch": 5.851900393184797, + "grad_norm": 0.4366179406642914, + "learning_rate": 2.1939076875883448e-05, + "loss": 0.3328, + "num_input_tokens_seen": 5985072, + "step": 22325 + }, + { + "epoch": 5.853211009174312, + "grad_norm": 1.7748178243637085, + "learning_rate": 2.1927725965197074e-05, + "loss": 0.4737, + "num_input_tokens_seen": 5986368, + "step": 22330 + }, + { + "epoch": 5.854521625163827, + "grad_norm": 1.009324312210083, + "learning_rate": 2.191637569753235e-05, + "loss": 0.2781, + "num_input_tokens_seen": 5987264, + "step": 22335 + }, + { + "epoch": 5.855832241153342, + "grad_norm": 2.1017491817474365, + "learning_rate": 2.1905026075264873e-05, + "loss": 0.2499, + "num_input_tokens_seen": 5988320, + "step": 22340 + }, + { + "epoch": 5.857142857142857, + "grad_norm": 1.1162234544754028, + "learning_rate": 2.1893677100770084e-05, + "loss": 0.2985, + "num_input_tokens_seen": 5989600, + "step": 22345 + }, + { + "epoch": 5.858453473132372, + "grad_norm": 1.786533236503601, + "learning_rate": 2.18823287764233e-05, + "loss": 0.4288, + "num_input_tokens_seen": 5991280, + "step": 22350 + }, + { + "epoch": 5.859764089121887, + "grad_norm": 1.5440572500228882, + "learning_rate": 2.187098110459973e-05, + "loss": 0.2603, + "num_input_tokens_seen": 5992416, + "step": 22355 + }, + { + "epoch": 5.8610747051114025, + "grad_norm": 1.3799378871917725, + "learning_rate": 2.1859634087674398e-05, + "loss": 0.312, + "num_input_tokens_seen": 5993680, + "step": 22360 + }, + { + "epoch": 5.862385321100917, + "grad_norm": 1.6058435440063477, + "learning_rate": 2.1848287728022233e-05, + "loss": 0.2789, + "num_input_tokens_seen": 5994912, + "step": 22365 + }, + { + "epoch": 5.863695937090433, + "grad_norm": 2.8585643768310547, + "learning_rate": 2.1836942028018e-05, + "loss": 0.3402, + "num_input_tokens_seen": 5996336, + "step": 22370 + }, + { + "epoch": 5.865006553079947, + "grad_norm": 1.5507372617721558, + "learning_rate": 2.182559699003634e-05, + "loss": 0.3084, + "num_input_tokens_seen": 5997504, + "step": 22375 + }, + { + "epoch": 5.866317169069463, + "grad_norm": 0.7258848547935486, + "learning_rate": 2.1814252616451735e-05, + "loss": 0.3369, + "num_input_tokens_seen": 5999040, + "step": 22380 + }, + { + "epoch": 5.8676277850589775, + "grad_norm": 1.480770230293274, + "learning_rate": 2.1802908909638558e-05, + "loss": 0.2533, + "num_input_tokens_seen": 6000336, + "step": 22385 + }, + { + "epoch": 5.868938401048493, + "grad_norm": 1.1923307180404663, + "learning_rate": 2.1791565871971024e-05, + "loss": 0.2464, + "num_input_tokens_seen": 6001584, + "step": 22390 + }, + { + "epoch": 5.870249017038008, + "grad_norm": 2.1142332553863525, + "learning_rate": 2.1780223505823212e-05, + "loss": 0.2989, + "num_input_tokens_seen": 6002688, + "step": 22395 + }, + { + "epoch": 5.871559633027523, + "grad_norm": 1.3939586877822876, + "learning_rate": 2.1768881813569046e-05, + "loss": 0.4203, + "num_input_tokens_seen": 6003952, + "step": 22400 + }, + { + "epoch": 5.872870249017038, + "grad_norm": 1.715328574180603, + "learning_rate": 2.175754079758234e-05, + "loss": 0.3378, + "num_input_tokens_seen": 6005184, + "step": 22405 + }, + { + "epoch": 5.8741808650065535, + "grad_norm": 1.2883825302124023, + "learning_rate": 2.1746200460236736e-05, + "loss": 0.3275, + "num_input_tokens_seen": 6006416, + "step": 22410 + }, + { + "epoch": 5.875491480996068, + "grad_norm": 1.1914499998092651, + "learning_rate": 2.173486080390576e-05, + "loss": 0.3323, + "num_input_tokens_seen": 6007616, + "step": 22415 + }, + { + "epoch": 5.876802096985584, + "grad_norm": 1.960267424583435, + "learning_rate": 2.1723521830962774e-05, + "loss": 0.3147, + "num_input_tokens_seen": 6009040, + "step": 22420 + }, + { + "epoch": 5.878112712975098, + "grad_norm": 0.768684983253479, + "learning_rate": 2.1712183543781005e-05, + "loss": 0.2423, + "num_input_tokens_seen": 6010144, + "step": 22425 + }, + { + "epoch": 5.879423328964613, + "grad_norm": 1.6042349338531494, + "learning_rate": 2.1700845944733543e-05, + "loss": 0.2744, + "num_input_tokens_seen": 6011232, + "step": 22430 + }, + { + "epoch": 5.8807339449541285, + "grad_norm": 1.6991816759109497, + "learning_rate": 2.1689509036193333e-05, + "loss": 0.263, + "num_input_tokens_seen": 6012096, + "step": 22435 + }, + { + "epoch": 5.882044560943643, + "grad_norm": 1.6937251091003418, + "learning_rate": 2.167817282053315e-05, + "loss": 0.4102, + "num_input_tokens_seen": 6013632, + "step": 22440 + }, + { + "epoch": 5.883355176933159, + "grad_norm": 3.579115867614746, + "learning_rate": 2.166683730012566e-05, + "loss": 0.354, + "num_input_tokens_seen": 6014608, + "step": 22445 + }, + { + "epoch": 5.884665792922673, + "grad_norm": 0.8665921688079834, + "learning_rate": 2.1655502477343363e-05, + "loss": 0.2816, + "num_input_tokens_seen": 6015824, + "step": 22450 + }, + { + "epoch": 5.885976408912189, + "grad_norm": 1.9726531505584717, + "learning_rate": 2.164416835455862e-05, + "loss": 0.3447, + "num_input_tokens_seen": 6017440, + "step": 22455 + }, + { + "epoch": 5.8872870249017035, + "grad_norm": 2.0385868549346924, + "learning_rate": 2.1632834934143642e-05, + "loss": 0.2427, + "num_input_tokens_seen": 6018592, + "step": 22460 + }, + { + "epoch": 5.888597640891219, + "grad_norm": 1.0874336957931519, + "learning_rate": 2.162150221847048e-05, + "loss": 0.2509, + "num_input_tokens_seen": 6019776, + "step": 22465 + }, + { + "epoch": 5.889908256880734, + "grad_norm": 0.7428236603736877, + "learning_rate": 2.1610170209911085e-05, + "loss": 0.4204, + "num_input_tokens_seen": 6021056, + "step": 22470 + }, + { + "epoch": 5.891218872870249, + "grad_norm": 1.5970890522003174, + "learning_rate": 2.15988389108372e-05, + "loss": 0.2983, + "num_input_tokens_seen": 6022528, + "step": 22475 + }, + { + "epoch": 5.892529488859764, + "grad_norm": 2.043182849884033, + "learning_rate": 2.1587508323620452e-05, + "loss": 0.3481, + "num_input_tokens_seen": 6023872, + "step": 22480 + }, + { + "epoch": 5.8938401048492794, + "grad_norm": 0.8339855074882507, + "learning_rate": 2.1576178450632324e-05, + "loss": 0.2839, + "num_input_tokens_seen": 6025600, + "step": 22485 + }, + { + "epoch": 5.895150720838794, + "grad_norm": 1.0956045389175415, + "learning_rate": 2.1564849294244126e-05, + "loss": 0.2619, + "num_input_tokens_seen": 6026864, + "step": 22490 + }, + { + "epoch": 5.89646133682831, + "grad_norm": 0.5216571092605591, + "learning_rate": 2.1553520856827038e-05, + "loss": 0.3217, + "num_input_tokens_seen": 6028016, + "step": 22495 + }, + { + "epoch": 5.897771952817824, + "grad_norm": 1.428098201751709, + "learning_rate": 2.1542193140752074e-05, + "loss": 0.3202, + "num_input_tokens_seen": 6029584, + "step": 22500 + }, + { + "epoch": 5.89908256880734, + "grad_norm": 1.773263931274414, + "learning_rate": 2.1530866148390114e-05, + "loss": 0.3385, + "num_input_tokens_seen": 6030640, + "step": 22505 + }, + { + "epoch": 5.9003931847968545, + "grad_norm": 1.9487859010696411, + "learning_rate": 2.151953988211187e-05, + "loss": 0.3568, + "num_input_tokens_seen": 6031904, + "step": 22510 + }, + { + "epoch": 5.90170380078637, + "grad_norm": 0.8840709328651428, + "learning_rate": 2.150821434428793e-05, + "loss": 0.3018, + "num_input_tokens_seen": 6033584, + "step": 22515 + }, + { + "epoch": 5.903014416775885, + "grad_norm": 1.8936331272125244, + "learning_rate": 2.149688953728867e-05, + "loss": 0.2466, + "num_input_tokens_seen": 6034704, + "step": 22520 + }, + { + "epoch": 5.904325032765399, + "grad_norm": 2.083932638168335, + "learning_rate": 2.1485565463484395e-05, + "loss": 0.5025, + "num_input_tokens_seen": 6036624, + "step": 22525 + }, + { + "epoch": 5.905635648754915, + "grad_norm": 1.2625293731689453, + "learning_rate": 2.1474242125245192e-05, + "loss": 0.3513, + "num_input_tokens_seen": 6037840, + "step": 22530 + }, + { + "epoch": 5.9069462647444295, + "grad_norm": 1.620772361755371, + "learning_rate": 2.1462919524941018e-05, + "loss": 0.3775, + "num_input_tokens_seen": 6039344, + "step": 22535 + }, + { + "epoch": 5.908256880733945, + "grad_norm": 1.2037286758422852, + "learning_rate": 2.145159766494168e-05, + "loss": 0.383, + "num_input_tokens_seen": 6040912, + "step": 22540 + }, + { + "epoch": 5.90956749672346, + "grad_norm": 1.5837451219558716, + "learning_rate": 2.1440276547616815e-05, + "loss": 0.2177, + "num_input_tokens_seen": 6042048, + "step": 22545 + }, + { + "epoch": 5.910878112712975, + "grad_norm": 1.117812991142273, + "learning_rate": 2.1428956175335922e-05, + "loss": 0.3166, + "num_input_tokens_seen": 6043328, + "step": 22550 + }, + { + "epoch": 5.91218872870249, + "grad_norm": 1.5233110189437866, + "learning_rate": 2.141763655046833e-05, + "loss": 0.2565, + "num_input_tokens_seen": 6044288, + "step": 22555 + }, + { + "epoch": 5.913499344692005, + "grad_norm": 1.271247148513794, + "learning_rate": 2.140631767538322e-05, + "loss": 0.2358, + "num_input_tokens_seen": 6045392, + "step": 22560 + }, + { + "epoch": 5.91480996068152, + "grad_norm": 1.0094654560089111, + "learning_rate": 2.1394999552449608e-05, + "loss": 0.3135, + "num_input_tokens_seen": 6046688, + "step": 22565 + }, + { + "epoch": 5.916120576671036, + "grad_norm": 2.1093404293060303, + "learning_rate": 2.1383682184036357e-05, + "loss": 0.4586, + "num_input_tokens_seen": 6047584, + "step": 22570 + }, + { + "epoch": 5.91743119266055, + "grad_norm": 14.251256942749023, + "learning_rate": 2.1372365572512188e-05, + "loss": 0.3766, + "num_input_tokens_seen": 6048912, + "step": 22575 + }, + { + "epoch": 5.918741808650066, + "grad_norm": 3.1968836784362793, + "learning_rate": 2.136104972024562e-05, + "loss": 0.2742, + "num_input_tokens_seen": 6049856, + "step": 22580 + }, + { + "epoch": 5.9200524246395805, + "grad_norm": 2.6276166439056396, + "learning_rate": 2.134973462960507e-05, + "loss": 0.2654, + "num_input_tokens_seen": 6051152, + "step": 22585 + }, + { + "epoch": 5.921363040629096, + "grad_norm": 1.1601897478103638, + "learning_rate": 2.133842030295875e-05, + "loss": 0.2205, + "num_input_tokens_seen": 6052672, + "step": 22590 + }, + { + "epoch": 5.922673656618611, + "grad_norm": 1.2493377923965454, + "learning_rate": 2.1327106742674735e-05, + "loss": 0.3506, + "num_input_tokens_seen": 6053840, + "step": 22595 + }, + { + "epoch": 5.923984272608125, + "grad_norm": 1.3348795175552368, + "learning_rate": 2.1315793951120934e-05, + "loss": 0.3251, + "num_input_tokens_seen": 6055696, + "step": 22600 + }, + { + "epoch": 5.925294888597641, + "grad_norm": 1.3208751678466797, + "learning_rate": 2.1304481930665095e-05, + "loss": 0.3285, + "num_input_tokens_seen": 6056720, + "step": 22605 + }, + { + "epoch": 5.926605504587156, + "grad_norm": 2.4137346744537354, + "learning_rate": 2.1293170683674803e-05, + "loss": 0.3336, + "num_input_tokens_seen": 6058080, + "step": 22610 + }, + { + "epoch": 5.927916120576671, + "grad_norm": 1.3559993505477905, + "learning_rate": 2.1281860212517477e-05, + "loss": 0.377, + "num_input_tokens_seen": 6059616, + "step": 22615 + }, + { + "epoch": 5.929226736566186, + "grad_norm": 0.9203824996948242, + "learning_rate": 2.1270550519560387e-05, + "loss": 0.3045, + "num_input_tokens_seen": 6061408, + "step": 22620 + }, + { + "epoch": 5.930537352555701, + "grad_norm": 0.9101852774620056, + "learning_rate": 2.125924160717063e-05, + "loss": 0.2191, + "num_input_tokens_seen": 6062864, + "step": 22625 + }, + { + "epoch": 5.931847968545216, + "grad_norm": 1.111655354499817, + "learning_rate": 2.1247933477715145e-05, + "loss": 0.2838, + "num_input_tokens_seen": 6064784, + "step": 22630 + }, + { + "epoch": 5.933158584534731, + "grad_norm": 0.9193039536476135, + "learning_rate": 2.123662613356068e-05, + "loss": 0.2735, + "num_input_tokens_seen": 6066464, + "step": 22635 + }, + { + "epoch": 5.934469200524246, + "grad_norm": 1.72642982006073, + "learning_rate": 2.1225319577073877e-05, + "loss": 0.2632, + "num_input_tokens_seen": 6067584, + "step": 22640 + }, + { + "epoch": 5.935779816513762, + "grad_norm": 0.6320211887359619, + "learning_rate": 2.121401381062117e-05, + "loss": 0.368, + "num_input_tokens_seen": 6068672, + "step": 22645 + }, + { + "epoch": 5.937090432503276, + "grad_norm": 0.7079176306724548, + "learning_rate": 2.1202708836568826e-05, + "loss": 0.2221, + "num_input_tokens_seen": 6070352, + "step": 22650 + }, + { + "epoch": 5.938401048492792, + "grad_norm": 0.9176162481307983, + "learning_rate": 2.1191404657282957e-05, + "loss": 0.3385, + "num_input_tokens_seen": 6071648, + "step": 22655 + }, + { + "epoch": 5.9397116644823065, + "grad_norm": 1.9438170194625854, + "learning_rate": 2.1180101275129515e-05, + "loss": 0.2169, + "num_input_tokens_seen": 6072528, + "step": 22660 + }, + { + "epoch": 5.941022280471822, + "grad_norm": 1.0929583311080933, + "learning_rate": 2.1168798692474277e-05, + "loss": 0.2464, + "num_input_tokens_seen": 6073744, + "step": 22665 + }, + { + "epoch": 5.942332896461337, + "grad_norm": 1.552949070930481, + "learning_rate": 2.115749691168286e-05, + "loss": 0.2838, + "num_input_tokens_seen": 6074848, + "step": 22670 + }, + { + "epoch": 5.943643512450852, + "grad_norm": 1.1311354637145996, + "learning_rate": 2.114619593512069e-05, + "loss": 0.2658, + "num_input_tokens_seen": 6076128, + "step": 22675 + }, + { + "epoch": 5.944954128440367, + "grad_norm": 1.4101852178573608, + "learning_rate": 2.1134895765153057e-05, + "loss": 0.2399, + "num_input_tokens_seen": 6077280, + "step": 22680 + }, + { + "epoch": 5.946264744429882, + "grad_norm": 1.7638297080993652, + "learning_rate": 2.112359640414506e-05, + "loss": 0.3779, + "num_input_tokens_seen": 6078768, + "step": 22685 + }, + { + "epoch": 5.947575360419397, + "grad_norm": 1.0809508562088013, + "learning_rate": 2.111229785446164e-05, + "loss": 0.352, + "num_input_tokens_seen": 6080528, + "step": 22690 + }, + { + "epoch": 5.948885976408912, + "grad_norm": 4.853146076202393, + "learning_rate": 2.110100011846755e-05, + "loss": 0.3598, + "num_input_tokens_seen": 6081792, + "step": 22695 + }, + { + "epoch": 5.950196592398427, + "grad_norm": 1.324357032775879, + "learning_rate": 2.1089703198527407e-05, + "loss": 0.2017, + "num_input_tokens_seen": 6083136, + "step": 22700 + }, + { + "epoch": 5.951507208387943, + "grad_norm": 1.502000093460083, + "learning_rate": 2.1078407097005632e-05, + "loss": 0.2815, + "num_input_tokens_seen": 6084240, + "step": 22705 + }, + { + "epoch": 5.952817824377457, + "grad_norm": 1.000220775604248, + "learning_rate": 2.106711181626647e-05, + "loss": 0.3472, + "num_input_tokens_seen": 6086384, + "step": 22710 + }, + { + "epoch": 5.954128440366972, + "grad_norm": 1.7664905786514282, + "learning_rate": 2.105581735867401e-05, + "loss": 0.3313, + "num_input_tokens_seen": 6087456, + "step": 22715 + }, + { + "epoch": 5.955439056356488, + "grad_norm": 1.3328217267990112, + "learning_rate": 2.1044523726592165e-05, + "loss": 0.2307, + "num_input_tokens_seen": 6088464, + "step": 22720 + }, + { + "epoch": 5.956749672346002, + "grad_norm": 1.6131354570388794, + "learning_rate": 2.1033230922384672e-05, + "loss": 0.2865, + "num_input_tokens_seen": 6090464, + "step": 22725 + }, + { + "epoch": 5.958060288335518, + "grad_norm": 3.1949820518493652, + "learning_rate": 2.1021938948415086e-05, + "loss": 0.2249, + "num_input_tokens_seen": 6091360, + "step": 22730 + }, + { + "epoch": 5.9593709043250325, + "grad_norm": 2.199178457260132, + "learning_rate": 2.1010647807046806e-05, + "loss": 0.2436, + "num_input_tokens_seen": 6092912, + "step": 22735 + }, + { + "epoch": 5.960681520314548, + "grad_norm": 3.0814547538757324, + "learning_rate": 2.099935750064305e-05, + "loss": 0.3265, + "num_input_tokens_seen": 6094384, + "step": 22740 + }, + { + "epoch": 5.961992136304063, + "grad_norm": 1.2987922430038452, + "learning_rate": 2.0988068031566854e-05, + "loss": 0.3902, + "num_input_tokens_seen": 6095616, + "step": 22745 + }, + { + "epoch": 5.963302752293578, + "grad_norm": 2.514639139175415, + "learning_rate": 2.097677940218108e-05, + "loss": 0.2906, + "num_input_tokens_seen": 6096720, + "step": 22750 + }, + { + "epoch": 5.964613368283093, + "grad_norm": 1.4615446329116821, + "learning_rate": 2.0965491614848433e-05, + "loss": 0.4488, + "num_input_tokens_seen": 6098592, + "step": 22755 + }, + { + "epoch": 5.965923984272608, + "grad_norm": 1.3620446920394897, + "learning_rate": 2.095420467193142e-05, + "loss": 0.3808, + "num_input_tokens_seen": 6099920, + "step": 22760 + }, + { + "epoch": 5.967234600262123, + "grad_norm": 1.1500799655914307, + "learning_rate": 2.0942918575792382e-05, + "loss": 0.4014, + "num_input_tokens_seen": 6101232, + "step": 22765 + }, + { + "epoch": 5.968545216251639, + "grad_norm": 1.3034989833831787, + "learning_rate": 2.0931633328793472e-05, + "loss": 0.2717, + "num_input_tokens_seen": 6102608, + "step": 22770 + }, + { + "epoch": 5.969855832241153, + "grad_norm": 1.1602181196212769, + "learning_rate": 2.0920348933296678e-05, + "loss": 0.2786, + "num_input_tokens_seen": 6104928, + "step": 22775 + }, + { + "epoch": 5.971166448230669, + "grad_norm": 1.82890784740448, + "learning_rate": 2.09090653916638e-05, + "loss": 0.4099, + "num_input_tokens_seen": 6106096, + "step": 22780 + }, + { + "epoch": 5.972477064220183, + "grad_norm": 1.8477575778961182, + "learning_rate": 2.0897782706256474e-05, + "loss": 0.2855, + "num_input_tokens_seen": 6107584, + "step": 22785 + }, + { + "epoch": 5.973787680209698, + "grad_norm": 1.3054826259613037, + "learning_rate": 2.0886500879436136e-05, + "loss": 0.3886, + "num_input_tokens_seen": 6109232, + "step": 22790 + }, + { + "epoch": 5.975098296199214, + "grad_norm": 1.151100754737854, + "learning_rate": 2.0875219913564053e-05, + "loss": 0.3223, + "num_input_tokens_seen": 6110384, + "step": 22795 + }, + { + "epoch": 5.976408912188729, + "grad_norm": 0.8557514548301697, + "learning_rate": 2.0863939811001316e-05, + "loss": 0.2494, + "num_input_tokens_seen": 6111840, + "step": 22800 + }, + { + "epoch": 5.977719528178244, + "grad_norm": 2.320065498352051, + "learning_rate": 2.0852660574108833e-05, + "loss": 0.2983, + "num_input_tokens_seen": 6113504, + "step": 22805 + }, + { + "epoch": 5.9790301441677585, + "grad_norm": 1.3823400735855103, + "learning_rate": 2.084138220524732e-05, + "loss": 0.233, + "num_input_tokens_seen": 6114576, + "step": 22810 + }, + { + "epoch": 5.980340760157274, + "grad_norm": 1.538938283920288, + "learning_rate": 2.0830104706777325e-05, + "loss": 0.289, + "num_input_tokens_seen": 6115632, + "step": 22815 + }, + { + "epoch": 5.981651376146789, + "grad_norm": 0.8550890684127808, + "learning_rate": 2.081882808105922e-05, + "loss": 0.2599, + "num_input_tokens_seen": 6117200, + "step": 22820 + }, + { + "epoch": 5.982961992136304, + "grad_norm": 1.4601541757583618, + "learning_rate": 2.0807552330453166e-05, + "loss": 0.2292, + "num_input_tokens_seen": 6118256, + "step": 22825 + }, + { + "epoch": 5.984272608125819, + "grad_norm": 0.9649074077606201, + "learning_rate": 2.079627745731917e-05, + "loss": 0.345, + "num_input_tokens_seen": 6119840, + "step": 22830 + }, + { + "epoch": 5.985583224115334, + "grad_norm": 0.9649481177330017, + "learning_rate": 2.0785003464017032e-05, + "loss": 0.2819, + "num_input_tokens_seen": 6121280, + "step": 22835 + }, + { + "epoch": 5.986893840104849, + "grad_norm": 0.7111610770225525, + "learning_rate": 2.0773730352906394e-05, + "loss": 0.2161, + "num_input_tokens_seen": 6122624, + "step": 22840 + }, + { + "epoch": 5.988204456094365, + "grad_norm": 1.1687846183776855, + "learning_rate": 2.076245812634669e-05, + "loss": 0.2645, + "num_input_tokens_seen": 6123504, + "step": 22845 + }, + { + "epoch": 5.989515072083879, + "grad_norm": 0.32466891407966614, + "learning_rate": 2.0751186786697175e-05, + "loss": 0.2769, + "num_input_tokens_seen": 6124496, + "step": 22850 + }, + { + "epoch": 5.990825688073395, + "grad_norm": 2.5452163219451904, + "learning_rate": 2.0739916336316924e-05, + "loss": 0.397, + "num_input_tokens_seen": 6125888, + "step": 22855 + }, + { + "epoch": 5.992136304062909, + "grad_norm": 1.3009552955627441, + "learning_rate": 2.072864677756482e-05, + "loss": 0.3974, + "num_input_tokens_seen": 6127520, + "step": 22860 + }, + { + "epoch": 5.993446920052425, + "grad_norm": 1.629966378211975, + "learning_rate": 2.0717378112799563e-05, + "loss": 0.2947, + "num_input_tokens_seen": 6128992, + "step": 22865 + }, + { + "epoch": 5.99475753604194, + "grad_norm": 2.202294111251831, + "learning_rate": 2.070611034437967e-05, + "loss": 0.2927, + "num_input_tokens_seen": 6130128, + "step": 22870 + }, + { + "epoch": 5.996068152031455, + "grad_norm": 2.1237881183624268, + "learning_rate": 2.0694843474663465e-05, + "loss": 0.2242, + "num_input_tokens_seen": 6131616, + "step": 22875 + }, + { + "epoch": 5.99737876802097, + "grad_norm": 1.1704553365707397, + "learning_rate": 2.0683577506009087e-05, + "loss": 0.3423, + "num_input_tokens_seen": 6133120, + "step": 22880 + }, + { + "epoch": 5.9986893840104845, + "grad_norm": 1.6856058835983276, + "learning_rate": 2.0672312440774473e-05, + "loss": 0.1497, + "num_input_tokens_seen": 6134224, + "step": 22885 + }, + { + "epoch": 6.0, + "grad_norm": 2.7754788398742676, + "learning_rate": 2.0661048281317388e-05, + "loss": 0.3128, + "num_input_tokens_seen": 6135704, + "step": 22890 + }, + { + "epoch": 6.001310615989515, + "grad_norm": 0.8929675817489624, + "learning_rate": 2.06497850299954e-05, + "loss": 0.2143, + "num_input_tokens_seen": 6137080, + "step": 22895 + }, + { + "epoch": 6.001572739187418, + "eval_loss": 0.5941077470779419, + "eval_runtime": 16.8744, + "eval_samples_per_second": 50.254, + "eval_steps_per_second": 25.127, + "num_input_tokens_seen": 6137320, + "step": 22896 + }, + { + "epoch": 6.00262123197903, + "grad_norm": 1.283185362815857, + "learning_rate": 2.063852268916589e-05, + "loss": 0.2131, + "num_input_tokens_seen": 6138248, + "step": 22900 + }, + { + "epoch": 6.003931847968545, + "grad_norm": 0.968062698841095, + "learning_rate": 2.062726126118605e-05, + "loss": 0.2009, + "num_input_tokens_seen": 6139960, + "step": 22905 + }, + { + "epoch": 6.00524246395806, + "grad_norm": 0.84856116771698, + "learning_rate": 2.0616000748412864e-05, + "loss": 0.2314, + "num_input_tokens_seen": 6141400, + "step": 22910 + }, + { + "epoch": 6.006553079947575, + "grad_norm": 2.1039388179779053, + "learning_rate": 2.060474115320315e-05, + "loss": 0.2027, + "num_input_tokens_seen": 6142536, + "step": 22915 + }, + { + "epoch": 6.007863695937091, + "grad_norm": 2.4904792308807373, + "learning_rate": 2.059348247791352e-05, + "loss": 0.3213, + "num_input_tokens_seen": 6143800, + "step": 22920 + }, + { + "epoch": 6.009174311926605, + "grad_norm": 2.230943202972412, + "learning_rate": 2.058222472490038e-05, + "loss": 0.2734, + "num_input_tokens_seen": 6144936, + "step": 22925 + }, + { + "epoch": 6.010484927916121, + "grad_norm": 1.1863491535186768, + "learning_rate": 2.0570967896519983e-05, + "loss": 0.2642, + "num_input_tokens_seen": 6146504, + "step": 22930 + }, + { + "epoch": 6.011795543905635, + "grad_norm": 1.1850225925445557, + "learning_rate": 2.0559711995128353e-05, + "loss": 0.2467, + "num_input_tokens_seen": 6147464, + "step": 22935 + }, + { + "epoch": 6.013106159895151, + "grad_norm": 1.3582134246826172, + "learning_rate": 2.0548457023081333e-05, + "loss": 0.1647, + "num_input_tokens_seen": 6148952, + "step": 22940 + }, + { + "epoch": 6.014416775884666, + "grad_norm": 1.0938668251037598, + "learning_rate": 2.0537202982734564e-05, + "loss": 0.1782, + "num_input_tokens_seen": 6150152, + "step": 22945 + }, + { + "epoch": 6.015727391874181, + "grad_norm": 1.959882378578186, + "learning_rate": 2.0525949876443497e-05, + "loss": 0.2718, + "num_input_tokens_seen": 6151592, + "step": 22950 + }, + { + "epoch": 6.017038007863696, + "grad_norm": 1.6970720291137695, + "learning_rate": 2.051469770656339e-05, + "loss": 0.2357, + "num_input_tokens_seen": 6152920, + "step": 22955 + }, + { + "epoch": 6.018348623853211, + "grad_norm": 2.386545419692993, + "learning_rate": 2.050344647544931e-05, + "loss": 0.3069, + "num_input_tokens_seen": 6154696, + "step": 22960 + }, + { + "epoch": 6.019659239842726, + "grad_norm": 2.083965301513672, + "learning_rate": 2.0492196185456108e-05, + "loss": 0.2869, + "num_input_tokens_seen": 6156200, + "step": 22965 + }, + { + "epoch": 6.0209698558322415, + "grad_norm": 1.5517297983169556, + "learning_rate": 2.0480946838938455e-05, + "loss": 0.2128, + "num_input_tokens_seen": 6157496, + "step": 22970 + }, + { + "epoch": 6.022280471821756, + "grad_norm": 4.422272205352783, + "learning_rate": 2.046969843825082e-05, + "loss": 0.2268, + "num_input_tokens_seen": 6158584, + "step": 22975 + }, + { + "epoch": 6.023591087811272, + "grad_norm": 2.417243003845215, + "learning_rate": 2.0458450985747468e-05, + "loss": 0.2273, + "num_input_tokens_seen": 6160120, + "step": 22980 + }, + { + "epoch": 6.024901703800786, + "grad_norm": 1.7893317937850952, + "learning_rate": 2.0447204483782482e-05, + "loss": 0.2509, + "num_input_tokens_seen": 6161240, + "step": 22985 + }, + { + "epoch": 6.026212319790301, + "grad_norm": 1.7639728784561157, + "learning_rate": 2.0435958934709738e-05, + "loss": 0.1686, + "num_input_tokens_seen": 6162408, + "step": 22990 + }, + { + "epoch": 6.027522935779817, + "grad_norm": 0.8152681589126587, + "learning_rate": 2.0424714340882893e-05, + "loss": 0.1392, + "num_input_tokens_seen": 6163704, + "step": 22995 + }, + { + "epoch": 6.028833551769331, + "grad_norm": 0.950090765953064, + "learning_rate": 2.0413470704655445e-05, + "loss": 0.2448, + "num_input_tokens_seen": 6164968, + "step": 23000 + }, + { + "epoch": 6.030144167758847, + "grad_norm": 0.9716110825538635, + "learning_rate": 2.0402228028380644e-05, + "loss": 0.2715, + "num_input_tokens_seen": 6168184, + "step": 23005 + }, + { + "epoch": 6.031454783748361, + "grad_norm": 0.47379323840141296, + "learning_rate": 2.0390986314411573e-05, + "loss": 0.0806, + "num_input_tokens_seen": 6169304, + "step": 23010 + }, + { + "epoch": 6.032765399737877, + "grad_norm": 0.19436359405517578, + "learning_rate": 2.03797455651011e-05, + "loss": 0.1949, + "num_input_tokens_seen": 6170328, + "step": 23015 + }, + { + "epoch": 6.034076015727392, + "grad_norm": 2.9670398235321045, + "learning_rate": 2.0368505782801904e-05, + "loss": 0.2034, + "num_input_tokens_seen": 6171112, + "step": 23020 + }, + { + "epoch": 6.035386631716907, + "grad_norm": 3.029914140701294, + "learning_rate": 2.0357266969866443e-05, + "loss": 0.2577, + "num_input_tokens_seen": 6172376, + "step": 23025 + }, + { + "epoch": 6.036697247706422, + "grad_norm": 3.062103033065796, + "learning_rate": 2.034602912864698e-05, + "loss": 0.2855, + "num_input_tokens_seen": 6173800, + "step": 23030 + }, + { + "epoch": 6.038007863695937, + "grad_norm": 2.1801652908325195, + "learning_rate": 2.0334792261495576e-05, + "loss": 0.2702, + "num_input_tokens_seen": 6175176, + "step": 23035 + }, + { + "epoch": 6.039318479685452, + "grad_norm": 3.1640803813934326, + "learning_rate": 2.032355637076409e-05, + "loss": 0.2479, + "num_input_tokens_seen": 6176856, + "step": 23040 + }, + { + "epoch": 6.0406290956749675, + "grad_norm": 1.7236089706420898, + "learning_rate": 2.031232145880418e-05, + "loss": 0.2433, + "num_input_tokens_seen": 6178072, + "step": 23045 + }, + { + "epoch": 6.041939711664482, + "grad_norm": 2.1499831676483154, + "learning_rate": 2.030108752796729e-05, + "loss": 0.23, + "num_input_tokens_seen": 6179224, + "step": 23050 + }, + { + "epoch": 6.043250327653998, + "grad_norm": 1.2278398275375366, + "learning_rate": 2.0289854580604667e-05, + "loss": 0.2068, + "num_input_tokens_seen": 6180312, + "step": 23055 + }, + { + "epoch": 6.044560943643512, + "grad_norm": 1.4195244312286377, + "learning_rate": 2.0278622619067334e-05, + "loss": 0.2516, + "num_input_tokens_seen": 6181896, + "step": 23060 + }, + { + "epoch": 6.045871559633028, + "grad_norm": 1.4342049360275269, + "learning_rate": 2.026739164570613e-05, + "loss": 0.2532, + "num_input_tokens_seen": 6183272, + "step": 23065 + }, + { + "epoch": 6.047182175622543, + "grad_norm": 1.7386977672576904, + "learning_rate": 2.0256161662871675e-05, + "loss": 0.2225, + "num_input_tokens_seen": 6184648, + "step": 23070 + }, + { + "epoch": 6.048492791612058, + "grad_norm": 2.5009238719940186, + "learning_rate": 2.0244932672914392e-05, + "loss": 0.1913, + "num_input_tokens_seen": 6185656, + "step": 23075 + }, + { + "epoch": 6.049803407601573, + "grad_norm": 2.2537295818328857, + "learning_rate": 2.0233704678184485e-05, + "loss": 0.1586, + "num_input_tokens_seen": 6186760, + "step": 23080 + }, + { + "epoch": 6.051114023591087, + "grad_norm": 1.5549808740615845, + "learning_rate": 2.0222477681031947e-05, + "loss": 0.2555, + "num_input_tokens_seen": 6188008, + "step": 23085 + }, + { + "epoch": 6.052424639580603, + "grad_norm": 2.49076509475708, + "learning_rate": 2.021125168380658e-05, + "loss": 0.2777, + "num_input_tokens_seen": 6189944, + "step": 23090 + }, + { + "epoch": 6.053735255570118, + "grad_norm": 0.5287485122680664, + "learning_rate": 2.0200026688857955e-05, + "loss": 0.1128, + "num_input_tokens_seen": 6190952, + "step": 23095 + }, + { + "epoch": 6.055045871559633, + "grad_norm": 1.0563832521438599, + "learning_rate": 2.0188802698535455e-05, + "loss": 0.2504, + "num_input_tokens_seen": 6192840, + "step": 23100 + }, + { + "epoch": 6.056356487549148, + "grad_norm": 1.7053354978561401, + "learning_rate": 2.0177579715188234e-05, + "loss": 0.1608, + "num_input_tokens_seen": 6193944, + "step": 23105 + }, + { + "epoch": 6.057667103538663, + "grad_norm": 1.4429682493209839, + "learning_rate": 2.016635774116525e-05, + "loss": 0.2891, + "num_input_tokens_seen": 6195368, + "step": 23110 + }, + { + "epoch": 6.058977719528178, + "grad_norm": 1.2314465045928955, + "learning_rate": 2.0155136778815246e-05, + "loss": 0.2633, + "num_input_tokens_seen": 6196536, + "step": 23115 + }, + { + "epoch": 6.0602883355176935, + "grad_norm": 1.7465370893478394, + "learning_rate": 2.0143916830486734e-05, + "loss": 0.2418, + "num_input_tokens_seen": 6197896, + "step": 23120 + }, + { + "epoch": 6.061598951507208, + "grad_norm": 1.3798727989196777, + "learning_rate": 2.0132697898528037e-05, + "loss": 0.2121, + "num_input_tokens_seen": 6199112, + "step": 23125 + }, + { + "epoch": 6.062909567496724, + "grad_norm": 1.5171958208084106, + "learning_rate": 2.0121479985287264e-05, + "loss": 0.2621, + "num_input_tokens_seen": 6200760, + "step": 23130 + }, + { + "epoch": 6.064220183486238, + "grad_norm": 2.7274653911590576, + "learning_rate": 2.01102630931123e-05, + "loss": 0.2228, + "num_input_tokens_seen": 6201816, + "step": 23135 + }, + { + "epoch": 6.065530799475754, + "grad_norm": 1.140273928642273, + "learning_rate": 2.0099047224350825e-05, + "loss": 0.3893, + "num_input_tokens_seen": 6203512, + "step": 23140 + }, + { + "epoch": 6.066841415465269, + "grad_norm": 1.3605625629425049, + "learning_rate": 2.0087832381350295e-05, + "loss": 0.2346, + "num_input_tokens_seen": 6204712, + "step": 23145 + }, + { + "epoch": 6.068152031454784, + "grad_norm": 2.3853139877319336, + "learning_rate": 2.007661856645796e-05, + "loss": 0.2198, + "num_input_tokens_seen": 6205640, + "step": 23150 + }, + { + "epoch": 6.069462647444299, + "grad_norm": 1.0734126567840576, + "learning_rate": 2.0065405782020844e-05, + "loss": 0.1677, + "num_input_tokens_seen": 6206520, + "step": 23155 + }, + { + "epoch": 6.070773263433814, + "grad_norm": 3.6580655574798584, + "learning_rate": 2.0054194030385774e-05, + "loss": 0.2687, + "num_input_tokens_seen": 6207656, + "step": 23160 + }, + { + "epoch": 6.072083879423329, + "grad_norm": 2.525531768798828, + "learning_rate": 2.0042983313899356e-05, + "loss": 0.2296, + "num_input_tokens_seen": 6208792, + "step": 23165 + }, + { + "epoch": 6.073394495412844, + "grad_norm": 2.1498658657073975, + "learning_rate": 2.0031773634907962e-05, + "loss": 0.2826, + "num_input_tokens_seen": 6210456, + "step": 23170 + }, + { + "epoch": 6.074705111402359, + "grad_norm": 0.4464334547519684, + "learning_rate": 2.0020564995757765e-05, + "loss": 0.1481, + "num_input_tokens_seen": 6211512, + "step": 23175 + }, + { + "epoch": 6.076015727391874, + "grad_norm": 1.1978397369384766, + "learning_rate": 2.0009357398794707e-05, + "loss": 0.2387, + "num_input_tokens_seen": 6213032, + "step": 23180 + }, + { + "epoch": 6.077326343381389, + "grad_norm": 1.0278284549713135, + "learning_rate": 1.999815084636452e-05, + "loss": 0.1905, + "num_input_tokens_seen": 6214152, + "step": 23185 + }, + { + "epoch": 6.078636959370904, + "grad_norm": 1.217346429824829, + "learning_rate": 1.9986945340812722e-05, + "loss": 0.2294, + "num_input_tokens_seen": 6215656, + "step": 23190 + }, + { + "epoch": 6.0799475753604195, + "grad_norm": 1.90694260597229, + "learning_rate": 1.9975740884484606e-05, + "loss": 0.3766, + "num_input_tokens_seen": 6216840, + "step": 23195 + }, + { + "epoch": 6.081258191349934, + "grad_norm": 1.1333776712417603, + "learning_rate": 1.9964537479725234e-05, + "loss": 0.2079, + "num_input_tokens_seen": 6218024, + "step": 23200 + }, + { + "epoch": 6.08256880733945, + "grad_norm": 3.2149181365966797, + "learning_rate": 1.9953335128879468e-05, + "loss": 0.2251, + "num_input_tokens_seen": 6219256, + "step": 23205 + }, + { + "epoch": 6.083879423328964, + "grad_norm": 1.175480842590332, + "learning_rate": 1.9942133834291936e-05, + "loss": 0.2764, + "num_input_tokens_seen": 6220872, + "step": 23210 + }, + { + "epoch": 6.08519003931848, + "grad_norm": 2.4639618396759033, + "learning_rate": 1.9930933598307058e-05, + "loss": 0.2013, + "num_input_tokens_seen": 6222216, + "step": 23215 + }, + { + "epoch": 6.086500655307995, + "grad_norm": 1.7261669635772705, + "learning_rate": 1.991973442326902e-05, + "loss": 0.1792, + "num_input_tokens_seen": 6223368, + "step": 23220 + }, + { + "epoch": 6.08781127129751, + "grad_norm": 2.1832637786865234, + "learning_rate": 1.990853631152179e-05, + "loss": 0.2297, + "num_input_tokens_seen": 6224696, + "step": 23225 + }, + { + "epoch": 6.089121887287025, + "grad_norm": 2.9617626667022705, + "learning_rate": 1.989733926540911e-05, + "loss": 0.226, + "num_input_tokens_seen": 6225768, + "step": 23230 + }, + { + "epoch": 6.09043250327654, + "grad_norm": 1.847436785697937, + "learning_rate": 1.988614328727451e-05, + "loss": 0.1735, + "num_input_tokens_seen": 6226808, + "step": 23235 + }, + { + "epoch": 6.091743119266055, + "grad_norm": 1.388616681098938, + "learning_rate": 1.9874948379461277e-05, + "loss": 0.3401, + "num_input_tokens_seen": 6228248, + "step": 23240 + }, + { + "epoch": 6.0930537352555705, + "grad_norm": 0.7159172892570496, + "learning_rate": 1.9863754544312495e-05, + "loss": 0.1551, + "num_input_tokens_seen": 6230104, + "step": 23245 + }, + { + "epoch": 6.094364351245085, + "grad_norm": 2.236185312271118, + "learning_rate": 1.9852561784171014e-05, + "loss": 0.1837, + "num_input_tokens_seen": 6231448, + "step": 23250 + }, + { + "epoch": 6.095674967234601, + "grad_norm": 1.6897507905960083, + "learning_rate": 1.9841370101379463e-05, + "loss": 0.2984, + "num_input_tokens_seen": 6232664, + "step": 23255 + }, + { + "epoch": 6.096985583224115, + "grad_norm": 0.8044047951698303, + "learning_rate": 1.983017949828023e-05, + "loss": 0.272, + "num_input_tokens_seen": 6234280, + "step": 23260 + }, + { + "epoch": 6.09829619921363, + "grad_norm": 1.7703752517700195, + "learning_rate": 1.981898997721548e-05, + "loss": 0.4531, + "num_input_tokens_seen": 6235528, + "step": 23265 + }, + { + "epoch": 6.0996068152031455, + "grad_norm": 2.5690009593963623, + "learning_rate": 1.98078015405272e-05, + "loss": 0.1953, + "num_input_tokens_seen": 6236536, + "step": 23270 + }, + { + "epoch": 6.10091743119266, + "grad_norm": 1.2277964353561401, + "learning_rate": 1.9796614190557074e-05, + "loss": 0.2953, + "num_input_tokens_seen": 6238088, + "step": 23275 + }, + { + "epoch": 6.102228047182176, + "grad_norm": 1.5945783853530884, + "learning_rate": 1.978542792964661e-05, + "loss": 0.242, + "num_input_tokens_seen": 6239336, + "step": 23280 + }, + { + "epoch": 6.10353866317169, + "grad_norm": 2.1003806591033936, + "learning_rate": 1.977424276013707e-05, + "loss": 0.278, + "num_input_tokens_seen": 6240680, + "step": 23285 + }, + { + "epoch": 6.104849279161206, + "grad_norm": 1.8079768419265747, + "learning_rate": 1.9763058684369495e-05, + "loss": 0.2068, + "num_input_tokens_seen": 6242296, + "step": 23290 + }, + { + "epoch": 6.1061598951507206, + "grad_norm": 0.6736743450164795, + "learning_rate": 1.9751875704684684e-05, + "loss": 0.1236, + "num_input_tokens_seen": 6243880, + "step": 23295 + }, + { + "epoch": 6.107470511140236, + "grad_norm": 2.7465505599975586, + "learning_rate": 1.974069382342322e-05, + "loss": 0.3905, + "num_input_tokens_seen": 6245160, + "step": 23300 + }, + { + "epoch": 6.108781127129751, + "grad_norm": 2.3322017192840576, + "learning_rate": 1.9729513042925455e-05, + "loss": 0.2123, + "num_input_tokens_seen": 6246280, + "step": 23305 + }, + { + "epoch": 6.110091743119266, + "grad_norm": 0.4060794413089752, + "learning_rate": 1.97183333655315e-05, + "loss": 0.1199, + "num_input_tokens_seen": 6247416, + "step": 23310 + }, + { + "epoch": 6.111402359108781, + "grad_norm": 1.3857381343841553, + "learning_rate": 1.970715479358125e-05, + "loss": 0.406, + "num_input_tokens_seen": 6248440, + "step": 23315 + }, + { + "epoch": 6.1127129750982965, + "grad_norm": 0.7591242790222168, + "learning_rate": 1.9695977329414352e-05, + "loss": 0.154, + "num_input_tokens_seen": 6249800, + "step": 23320 + }, + { + "epoch": 6.114023591087811, + "grad_norm": 1.3657106161117554, + "learning_rate": 1.9684800975370224e-05, + "loss": 0.1485, + "num_input_tokens_seen": 6250856, + "step": 23325 + }, + { + "epoch": 6.115334207077327, + "grad_norm": 1.0496255159378052, + "learning_rate": 1.9673625733788084e-05, + "loss": 0.2499, + "num_input_tokens_seen": 6252376, + "step": 23330 + }, + { + "epoch": 6.116644823066841, + "grad_norm": 3.7419278621673584, + "learning_rate": 1.966245160700687e-05, + "loss": 0.1415, + "num_input_tokens_seen": 6253592, + "step": 23335 + }, + { + "epoch": 6.117955439056357, + "grad_norm": 2.5733492374420166, + "learning_rate": 1.965127859736531e-05, + "loss": 0.1449, + "num_input_tokens_seen": 6255208, + "step": 23340 + }, + { + "epoch": 6.1192660550458715, + "grad_norm": 2.492828845977783, + "learning_rate": 1.96401067072019e-05, + "loss": 0.2409, + "num_input_tokens_seen": 6256216, + "step": 23345 + }, + { + "epoch": 6.120576671035387, + "grad_norm": 1.3550175428390503, + "learning_rate": 1.9628935938854897e-05, + "loss": 0.1805, + "num_input_tokens_seen": 6257752, + "step": 23350 + }, + { + "epoch": 6.121887287024902, + "grad_norm": 1.374930500984192, + "learning_rate": 1.961776629466232e-05, + "loss": 0.2447, + "num_input_tokens_seen": 6258904, + "step": 23355 + }, + { + "epoch": 6.123197903014416, + "grad_norm": 1.6510207653045654, + "learning_rate": 1.9606597776961955e-05, + "loss": 0.263, + "num_input_tokens_seen": 6260120, + "step": 23360 + }, + { + "epoch": 6.124508519003932, + "grad_norm": 1.7457332611083984, + "learning_rate": 1.9595430388091358e-05, + "loss": 0.1526, + "num_input_tokens_seen": 6261448, + "step": 23365 + }, + { + "epoch": 6.1258191349934465, + "grad_norm": 2.043644428253174, + "learning_rate": 1.958426413038784e-05, + "loss": 0.2601, + "num_input_tokens_seen": 6262728, + "step": 23370 + }, + { + "epoch": 6.127129750982962, + "grad_norm": 1.8224364519119263, + "learning_rate": 1.957309900618849e-05, + "loss": 0.2582, + "num_input_tokens_seen": 6264216, + "step": 23375 + }, + { + "epoch": 6.128440366972477, + "grad_norm": 1.7963765859603882, + "learning_rate": 1.956193501783012e-05, + "loss": 0.2313, + "num_input_tokens_seen": 6265464, + "step": 23380 + }, + { + "epoch": 6.129750982961992, + "grad_norm": 1.435902714729309, + "learning_rate": 1.9550772167649378e-05, + "loss": 0.128, + "num_input_tokens_seen": 6266680, + "step": 23385 + }, + { + "epoch": 6.131061598951507, + "grad_norm": 5.07275915145874, + "learning_rate": 1.9539610457982596e-05, + "loss": 0.4278, + "num_input_tokens_seen": 6267640, + "step": 23390 + }, + { + "epoch": 6.1323722149410225, + "grad_norm": 1.4183670282363892, + "learning_rate": 1.9528449891165913e-05, + "loss": 0.2789, + "num_input_tokens_seen": 6269080, + "step": 23395 + }, + { + "epoch": 6.133682830930537, + "grad_norm": 1.9014067649841309, + "learning_rate": 1.951729046953521e-05, + "loss": 0.2138, + "num_input_tokens_seen": 6270360, + "step": 23400 + }, + { + "epoch": 6.134993446920053, + "grad_norm": 2.522146463394165, + "learning_rate": 1.9506132195426138e-05, + "loss": 0.2762, + "num_input_tokens_seen": 6271720, + "step": 23405 + }, + { + "epoch": 6.136304062909567, + "grad_norm": 2.3423736095428467, + "learning_rate": 1.9494975071174112e-05, + "loss": 0.2735, + "num_input_tokens_seen": 6273352, + "step": 23410 + }, + { + "epoch": 6.137614678899083, + "grad_norm": 3.0029685497283936, + "learning_rate": 1.9483819099114286e-05, + "loss": 0.2626, + "num_input_tokens_seen": 6274808, + "step": 23415 + }, + { + "epoch": 6.1389252948885975, + "grad_norm": 1.879326581954956, + "learning_rate": 1.9472664281581594e-05, + "loss": 0.2521, + "num_input_tokens_seen": 6275704, + "step": 23420 + }, + { + "epoch": 6.140235910878113, + "grad_norm": 1.5438194274902344, + "learning_rate": 1.9461510620910713e-05, + "loss": 0.2331, + "num_input_tokens_seen": 6277400, + "step": 23425 + }, + { + "epoch": 6.141546526867628, + "grad_norm": 0.9421892166137695, + "learning_rate": 1.94503581194361e-05, + "loss": 0.2608, + "num_input_tokens_seen": 6278696, + "step": 23430 + }, + { + "epoch": 6.142857142857143, + "grad_norm": 2.7862508296966553, + "learning_rate": 1.9439206779491934e-05, + "loss": 0.2084, + "num_input_tokens_seen": 6280024, + "step": 23435 + }, + { + "epoch": 6.144167758846658, + "grad_norm": 1.9980436563491821, + "learning_rate": 1.942805660341217e-05, + "loss": 0.339, + "num_input_tokens_seen": 6281048, + "step": 23440 + }, + { + "epoch": 6.145478374836173, + "grad_norm": 0.7072227001190186, + "learning_rate": 1.9416907593530548e-05, + "loss": 0.1825, + "num_input_tokens_seen": 6282360, + "step": 23445 + }, + { + "epoch": 6.146788990825688, + "grad_norm": 3.549912452697754, + "learning_rate": 1.9405759752180515e-05, + "loss": 0.1212, + "num_input_tokens_seen": 6283080, + "step": 23450 + }, + { + "epoch": 6.148099606815203, + "grad_norm": 1.3201016187667847, + "learning_rate": 1.9394613081695302e-05, + "loss": 0.168, + "num_input_tokens_seen": 6284824, + "step": 23455 + }, + { + "epoch": 6.149410222804718, + "grad_norm": 4.181965351104736, + "learning_rate": 1.9383467584407883e-05, + "loss": 0.2038, + "num_input_tokens_seen": 6285896, + "step": 23460 + }, + { + "epoch": 6.150720838794233, + "grad_norm": 1.5191190242767334, + "learning_rate": 1.9372323262650992e-05, + "loss": 0.3275, + "num_input_tokens_seen": 6287080, + "step": 23465 + }, + { + "epoch": 6.1520314547837485, + "grad_norm": 2.590013027191162, + "learning_rate": 1.936118011875713e-05, + "loss": 0.2459, + "num_input_tokens_seen": 6288088, + "step": 23470 + }, + { + "epoch": 6.153342070773263, + "grad_norm": 1.963974952697754, + "learning_rate": 1.9350038155058515e-05, + "loss": 0.2402, + "num_input_tokens_seen": 6289176, + "step": 23475 + }, + { + "epoch": 6.154652686762779, + "grad_norm": 1.7785799503326416, + "learning_rate": 1.9338897373887154e-05, + "loss": 0.2596, + "num_input_tokens_seen": 6290376, + "step": 23480 + }, + { + "epoch": 6.155963302752293, + "grad_norm": 2.4544460773468018, + "learning_rate": 1.932775777757479e-05, + "loss": 0.3145, + "num_input_tokens_seen": 6291688, + "step": 23485 + }, + { + "epoch": 6.157273918741809, + "grad_norm": 1.0536154508590698, + "learning_rate": 1.9316619368452928e-05, + "loss": 0.1655, + "num_input_tokens_seen": 6292792, + "step": 23490 + }, + { + "epoch": 6.1585845347313235, + "grad_norm": 4.229962348937988, + "learning_rate": 1.9305482148852794e-05, + "loss": 0.2341, + "num_input_tokens_seen": 6294184, + "step": 23495 + }, + { + "epoch": 6.159895150720839, + "grad_norm": 4.158444881439209, + "learning_rate": 1.929434612110542e-05, + "loss": 0.1901, + "num_input_tokens_seen": 6295384, + "step": 23500 + }, + { + "epoch": 6.161205766710354, + "grad_norm": 3.414186716079712, + "learning_rate": 1.9283211287541548e-05, + "loss": 0.2422, + "num_input_tokens_seen": 6296568, + "step": 23505 + }, + { + "epoch": 6.162516382699869, + "grad_norm": 1.7470489740371704, + "learning_rate": 1.9272077650491673e-05, + "loss": 0.2354, + "num_input_tokens_seen": 6297880, + "step": 23510 + }, + { + "epoch": 6.163826998689384, + "grad_norm": 2.909095525741577, + "learning_rate": 1.9260945212286047e-05, + "loss": 0.3579, + "num_input_tokens_seen": 6298824, + "step": 23515 + }, + { + "epoch": 6.165137614678899, + "grad_norm": 1.6522356271743774, + "learning_rate": 1.9249813975254673e-05, + "loss": 0.156, + "num_input_tokens_seen": 6299960, + "step": 23520 + }, + { + "epoch": 6.166448230668414, + "grad_norm": 2.1124961376190186, + "learning_rate": 1.92386839417273e-05, + "loss": 0.2367, + "num_input_tokens_seen": 6301272, + "step": 23525 + }, + { + "epoch": 6.16775884665793, + "grad_norm": 2.4552671909332275, + "learning_rate": 1.9227555114033422e-05, + "loss": 0.3056, + "num_input_tokens_seen": 6302504, + "step": 23530 + }, + { + "epoch": 6.169069462647444, + "grad_norm": 2.0773024559020996, + "learning_rate": 1.921642749450228e-05, + "loss": 0.2398, + "num_input_tokens_seen": 6303832, + "step": 23535 + }, + { + "epoch": 6.17038007863696, + "grad_norm": 0.8869185447692871, + "learning_rate": 1.9205301085462877e-05, + "loss": 0.2267, + "num_input_tokens_seen": 6305336, + "step": 23540 + }, + { + "epoch": 6.1716906946264745, + "grad_norm": 2.230971574783325, + "learning_rate": 1.919417588924394e-05, + "loss": 0.3185, + "num_input_tokens_seen": 6306568, + "step": 23545 + }, + { + "epoch": 6.173001310615989, + "grad_norm": 1.4933359622955322, + "learning_rate": 1.9183051908173963e-05, + "loss": 0.2472, + "num_input_tokens_seen": 6307880, + "step": 23550 + }, + { + "epoch": 6.174311926605505, + "grad_norm": 0.6358409523963928, + "learning_rate": 1.9171929144581156e-05, + "loss": 0.2072, + "num_input_tokens_seen": 6309448, + "step": 23555 + }, + { + "epoch": 6.175622542595019, + "grad_norm": 1.0385595560073853, + "learning_rate": 1.9160807600793517e-05, + "loss": 0.2293, + "num_input_tokens_seen": 6310888, + "step": 23560 + }, + { + "epoch": 6.176933158584535, + "grad_norm": 2.252379894256592, + "learning_rate": 1.9149687279138768e-05, + "loss": 0.377, + "num_input_tokens_seen": 6312344, + "step": 23565 + }, + { + "epoch": 6.1782437745740495, + "grad_norm": 2.5613276958465576, + "learning_rate": 1.9138568181944357e-05, + "loss": 0.3616, + "num_input_tokens_seen": 6313704, + "step": 23570 + }, + { + "epoch": 6.179554390563565, + "grad_norm": 2.2154698371887207, + "learning_rate": 1.9127450311537494e-05, + "loss": 0.2368, + "num_input_tokens_seen": 6315016, + "step": 23575 + }, + { + "epoch": 6.18086500655308, + "grad_norm": 2.542289972305298, + "learning_rate": 1.9116333670245135e-05, + "loss": 0.1969, + "num_input_tokens_seen": 6315992, + "step": 23580 + }, + { + "epoch": 6.182175622542595, + "grad_norm": 2.525702714920044, + "learning_rate": 1.9105218260393975e-05, + "loss": 0.2694, + "num_input_tokens_seen": 6317320, + "step": 23585 + }, + { + "epoch": 6.18348623853211, + "grad_norm": 1.7039753198623657, + "learning_rate": 1.9094104084310446e-05, + "loss": 0.4943, + "num_input_tokens_seen": 6318424, + "step": 23590 + }, + { + "epoch": 6.184796854521625, + "grad_norm": 4.411831378936768, + "learning_rate": 1.908299114432073e-05, + "loss": 0.2474, + "num_input_tokens_seen": 6319736, + "step": 23595 + }, + { + "epoch": 6.18610747051114, + "grad_norm": 1.316947340965271, + "learning_rate": 1.907187944275074e-05, + "loss": 0.1775, + "num_input_tokens_seen": 6320904, + "step": 23600 + }, + { + "epoch": 6.187418086500656, + "grad_norm": 18.549659729003906, + "learning_rate": 1.9060768981926137e-05, + "loss": 0.3734, + "num_input_tokens_seen": 6321896, + "step": 23605 + }, + { + "epoch": 6.18872870249017, + "grad_norm": 2.3596088886260986, + "learning_rate": 1.904965976417232e-05, + "loss": 0.184, + "num_input_tokens_seen": 6323176, + "step": 23610 + }, + { + "epoch": 6.190039318479686, + "grad_norm": 1.9442670345306396, + "learning_rate": 1.9038551791814446e-05, + "loss": 0.2088, + "num_input_tokens_seen": 6324776, + "step": 23615 + }, + { + "epoch": 6.1913499344692005, + "grad_norm": 1.3402796983718872, + "learning_rate": 1.902744506717738e-05, + "loss": 0.2582, + "num_input_tokens_seen": 6326152, + "step": 23620 + }, + { + "epoch": 6.192660550458716, + "grad_norm": 1.2271294593811035, + "learning_rate": 1.9016339592585737e-05, + "loss": 0.1507, + "num_input_tokens_seen": 6327112, + "step": 23625 + }, + { + "epoch": 6.193971166448231, + "grad_norm": 1.0585834980010986, + "learning_rate": 1.9005235370363877e-05, + "loss": 0.3352, + "num_input_tokens_seen": 6328616, + "step": 23630 + }, + { + "epoch": 6.195281782437746, + "grad_norm": 2.7060294151306152, + "learning_rate": 1.89941324028359e-05, + "loss": 0.2145, + "num_input_tokens_seen": 6329736, + "step": 23635 + }, + { + "epoch": 6.196592398427261, + "grad_norm": 2.0520992279052734, + "learning_rate": 1.8983030692325632e-05, + "loss": 0.217, + "num_input_tokens_seen": 6330984, + "step": 23640 + }, + { + "epoch": 6.1979030144167755, + "grad_norm": 1.7032057046890259, + "learning_rate": 1.8971930241156646e-05, + "loss": 0.2225, + "num_input_tokens_seen": 6332168, + "step": 23645 + }, + { + "epoch": 6.199213630406291, + "grad_norm": 0.8283475041389465, + "learning_rate": 1.8960831051652243e-05, + "loss": 0.1661, + "num_input_tokens_seen": 6333320, + "step": 23650 + }, + { + "epoch": 6.200524246395806, + "grad_norm": 1.1808310747146606, + "learning_rate": 1.8949733126135464e-05, + "loss": 0.2249, + "num_input_tokens_seen": 6334712, + "step": 23655 + }, + { + "epoch": 6.201834862385321, + "grad_norm": 1.4049193859100342, + "learning_rate": 1.8938636466929088e-05, + "loss": 0.2928, + "num_input_tokens_seen": 6335752, + "step": 23660 + }, + { + "epoch": 6.203145478374836, + "grad_norm": 1.3506301641464233, + "learning_rate": 1.8927541076355626e-05, + "loss": 0.1857, + "num_input_tokens_seen": 6336872, + "step": 23665 + }, + { + "epoch": 6.204456094364351, + "grad_norm": 2.0092902183532715, + "learning_rate": 1.8916446956737317e-05, + "loss": 0.4782, + "num_input_tokens_seen": 6338296, + "step": 23670 + }, + { + "epoch": 6.205766710353866, + "grad_norm": 1.5817691087722778, + "learning_rate": 1.890535411039616e-05, + "loss": 0.2599, + "num_input_tokens_seen": 6339560, + "step": 23675 + }, + { + "epoch": 6.207077326343382, + "grad_norm": 1.5899996757507324, + "learning_rate": 1.8894262539653858e-05, + "loss": 0.1993, + "num_input_tokens_seen": 6340744, + "step": 23680 + }, + { + "epoch": 6.208387942332896, + "grad_norm": 1.335041880607605, + "learning_rate": 1.888317224683185e-05, + "loss": 0.2072, + "num_input_tokens_seen": 6342040, + "step": 23685 + }, + { + "epoch": 6.209698558322412, + "grad_norm": 1.546730875968933, + "learning_rate": 1.8872083234251324e-05, + "loss": 0.2162, + "num_input_tokens_seen": 6343000, + "step": 23690 + }, + { + "epoch": 6.2110091743119265, + "grad_norm": 1.469709873199463, + "learning_rate": 1.8860995504233188e-05, + "loss": 0.2386, + "num_input_tokens_seen": 6343992, + "step": 23695 + }, + { + "epoch": 6.212319790301442, + "grad_norm": 1.7224047183990479, + "learning_rate": 1.8849909059098087e-05, + "loss": 0.4022, + "num_input_tokens_seen": 6345640, + "step": 23700 + }, + { + "epoch": 6.213630406290957, + "grad_norm": 1.3010834455490112, + "learning_rate": 1.8838823901166396e-05, + "loss": 0.2033, + "num_input_tokens_seen": 6347128, + "step": 23705 + }, + { + "epoch": 6.214941022280472, + "grad_norm": 1.3797072172164917, + "learning_rate": 1.8827740032758215e-05, + "loss": 0.2805, + "num_input_tokens_seen": 6348872, + "step": 23710 + }, + { + "epoch": 6.216251638269987, + "grad_norm": 1.4479491710662842, + "learning_rate": 1.8816657456193376e-05, + "loss": 0.2333, + "num_input_tokens_seen": 6350536, + "step": 23715 + }, + { + "epoch": 6.217562254259502, + "grad_norm": 0.8714211583137512, + "learning_rate": 1.880557617379145e-05, + "loss": 0.1608, + "num_input_tokens_seen": 6352200, + "step": 23720 + }, + { + "epoch": 6.218872870249017, + "grad_norm": 4.09882926940918, + "learning_rate": 1.8794496187871725e-05, + "loss": 0.3079, + "num_input_tokens_seen": 6353528, + "step": 23725 + }, + { + "epoch": 6.220183486238533, + "grad_norm": 1.4358805418014526, + "learning_rate": 1.8783417500753224e-05, + "loss": 0.2708, + "num_input_tokens_seen": 6355016, + "step": 23730 + }, + { + "epoch": 6.221494102228047, + "grad_norm": 1.1673263311386108, + "learning_rate": 1.8772340114754705e-05, + "loss": 0.1723, + "num_input_tokens_seen": 6356536, + "step": 23735 + }, + { + "epoch": 6.222804718217562, + "grad_norm": 1.9292746782302856, + "learning_rate": 1.8761264032194642e-05, + "loss": 0.2158, + "num_input_tokens_seen": 6358200, + "step": 23740 + }, + { + "epoch": 6.224115334207077, + "grad_norm": 2.073211431503296, + "learning_rate": 1.875018925539123e-05, + "loss": 0.137, + "num_input_tokens_seen": 6359240, + "step": 23745 + }, + { + "epoch": 6.225425950196592, + "grad_norm": 1.802865982055664, + "learning_rate": 1.8739115786662415e-05, + "loss": 0.2209, + "num_input_tokens_seen": 6361096, + "step": 23750 + }, + { + "epoch": 6.226736566186108, + "grad_norm": 2.222688674926758, + "learning_rate": 1.872804362832584e-05, + "loss": 0.1755, + "num_input_tokens_seen": 6362200, + "step": 23755 + }, + { + "epoch": 6.228047182175622, + "grad_norm": 1.4231433868408203, + "learning_rate": 1.8716972782698905e-05, + "loss": 0.1273, + "num_input_tokens_seen": 6363448, + "step": 23760 + }, + { + "epoch": 6.229357798165138, + "grad_norm": 1.8141772747039795, + "learning_rate": 1.8705903252098704e-05, + "loss": 0.1686, + "num_input_tokens_seen": 6364376, + "step": 23765 + }, + { + "epoch": 6.2306684141546524, + "grad_norm": 1.0569891929626465, + "learning_rate": 1.869483503884208e-05, + "loss": 0.2758, + "num_input_tokens_seen": 6365688, + "step": 23770 + }, + { + "epoch": 6.231979030144168, + "grad_norm": 0.6782994866371155, + "learning_rate": 1.8683768145245586e-05, + "loss": 0.1727, + "num_input_tokens_seen": 6366808, + "step": 23775 + }, + { + "epoch": 6.233289646133683, + "grad_norm": 1.758559226989746, + "learning_rate": 1.8672702573625505e-05, + "loss": 0.3485, + "num_input_tokens_seen": 6368296, + "step": 23780 + }, + { + "epoch": 6.234600262123198, + "grad_norm": 0.5242366790771484, + "learning_rate": 1.866163832629784e-05, + "loss": 0.2934, + "num_input_tokens_seen": 6369224, + "step": 23785 + }, + { + "epoch": 6.235910878112713, + "grad_norm": 1.7976456880569458, + "learning_rate": 1.8650575405578324e-05, + "loss": 0.2129, + "num_input_tokens_seen": 6370232, + "step": 23790 + }, + { + "epoch": 6.237221494102228, + "grad_norm": 1.663989543914795, + "learning_rate": 1.8639513813782407e-05, + "loss": 0.236, + "num_input_tokens_seen": 6371336, + "step": 23795 + }, + { + "epoch": 6.238532110091743, + "grad_norm": 1.7920814752578735, + "learning_rate": 1.8628453553225262e-05, + "loss": 0.2761, + "num_input_tokens_seen": 6372824, + "step": 23800 + }, + { + "epoch": 6.239842726081259, + "grad_norm": 2.4470391273498535, + "learning_rate": 1.8617394626221774e-05, + "loss": 0.234, + "num_input_tokens_seen": 6374168, + "step": 23805 + }, + { + "epoch": 6.241153342070773, + "grad_norm": 1.2032989263534546, + "learning_rate": 1.8606337035086564e-05, + "loss": 0.2396, + "num_input_tokens_seen": 6375576, + "step": 23810 + }, + { + "epoch": 6.242463958060289, + "grad_norm": 0.9677314162254333, + "learning_rate": 1.859528078213396e-05, + "loss": 0.3501, + "num_input_tokens_seen": 6376584, + "step": 23815 + }, + { + "epoch": 6.243774574049803, + "grad_norm": 1.045444369316101, + "learning_rate": 1.8584225869678033e-05, + "loss": 0.238, + "num_input_tokens_seen": 6377752, + "step": 23820 + }, + { + "epoch": 6.245085190039318, + "grad_norm": 1.1063047647476196, + "learning_rate": 1.8573172300032538e-05, + "loss": 0.2709, + "num_input_tokens_seen": 6378776, + "step": 23825 + }, + { + "epoch": 6.246395806028834, + "grad_norm": 1.2190821170806885, + "learning_rate": 1.8562120075510976e-05, + "loss": 0.1619, + "num_input_tokens_seen": 6379976, + "step": 23830 + }, + { + "epoch": 6.247706422018348, + "grad_norm": 1.7504301071166992, + "learning_rate": 1.8551069198426556e-05, + "loss": 0.1666, + "num_input_tokens_seen": 6381768, + "step": 23835 + }, + { + "epoch": 6.249017038007864, + "grad_norm": 1.960657000541687, + "learning_rate": 1.854001967109221e-05, + "loss": 0.4169, + "num_input_tokens_seen": 6383256, + "step": 23840 + }, + { + "epoch": 6.250327653997378, + "grad_norm": 1.069773554801941, + "learning_rate": 1.8528971495820584e-05, + "loss": 0.2085, + "num_input_tokens_seen": 6384744, + "step": 23845 + }, + { + "epoch": 6.251638269986894, + "grad_norm": 2.8704209327697754, + "learning_rate": 1.8517924674924048e-05, + "loss": 0.2708, + "num_input_tokens_seen": 6386136, + "step": 23850 + }, + { + "epoch": 6.252948885976409, + "grad_norm": 1.4593420028686523, + "learning_rate": 1.8506879210714685e-05, + "loss": 0.1497, + "num_input_tokens_seen": 6387160, + "step": 23855 + }, + { + "epoch": 6.254259501965924, + "grad_norm": 0.7323416471481323, + "learning_rate": 1.849583510550427e-05, + "loss": 0.1553, + "num_input_tokens_seen": 6389128, + "step": 23860 + }, + { + "epoch": 6.255570117955439, + "grad_norm": 5.14224910736084, + "learning_rate": 1.8484792361604336e-05, + "loss": 0.219, + "num_input_tokens_seen": 6390120, + "step": 23865 + }, + { + "epoch": 6.256880733944954, + "grad_norm": 1.3043314218521118, + "learning_rate": 1.84737509813261e-05, + "loss": 0.205, + "num_input_tokens_seen": 6391544, + "step": 23870 + }, + { + "epoch": 6.258191349934469, + "grad_norm": 1.2266584634780884, + "learning_rate": 1.8462710966980506e-05, + "loss": 0.2279, + "num_input_tokens_seen": 6392792, + "step": 23875 + }, + { + "epoch": 6.259501965923985, + "grad_norm": 1.3236579895019531, + "learning_rate": 1.845167232087822e-05, + "loss": 0.2422, + "num_input_tokens_seen": 6394248, + "step": 23880 + }, + { + "epoch": 6.260812581913499, + "grad_norm": 1.092944622039795, + "learning_rate": 1.8440635045329595e-05, + "loss": 0.1791, + "num_input_tokens_seen": 6395416, + "step": 23885 + }, + { + "epoch": 6.262123197903015, + "grad_norm": 2.1513638496398926, + "learning_rate": 1.8429599142644724e-05, + "loss": 0.2051, + "num_input_tokens_seen": 6396536, + "step": 23890 + }, + { + "epoch": 6.263433813892529, + "grad_norm": 1.6296608448028564, + "learning_rate": 1.84185646151334e-05, + "loss": 0.2333, + "num_input_tokens_seen": 6397704, + "step": 23895 + }, + { + "epoch": 6.264744429882045, + "grad_norm": 3.908989667892456, + "learning_rate": 1.840753146510512e-05, + "loss": 0.2519, + "num_input_tokens_seen": 6398840, + "step": 23900 + }, + { + "epoch": 6.26605504587156, + "grad_norm": 1.0339435338974, + "learning_rate": 1.8396499694869123e-05, + "loss": 0.2565, + "num_input_tokens_seen": 6400264, + "step": 23905 + }, + { + "epoch": 6.267365661861074, + "grad_norm": 1.8261680603027344, + "learning_rate": 1.8385469306734327e-05, + "loss": 0.2272, + "num_input_tokens_seen": 6401224, + "step": 23910 + }, + { + "epoch": 6.26867627785059, + "grad_norm": 4.4918928146362305, + "learning_rate": 1.8374440303009382e-05, + "loss": 0.3154, + "num_input_tokens_seen": 6403080, + "step": 23915 + }, + { + "epoch": 6.269986893840104, + "grad_norm": 0.902696430683136, + "learning_rate": 1.836341268600263e-05, + "loss": 0.2004, + "num_input_tokens_seen": 6404280, + "step": 23920 + }, + { + "epoch": 6.27129750982962, + "grad_norm": 1.886946439743042, + "learning_rate": 1.8352386458022135e-05, + "loss": 0.2674, + "num_input_tokens_seen": 6405656, + "step": 23925 + }, + { + "epoch": 6.272608125819135, + "grad_norm": 2.3929038047790527, + "learning_rate": 1.8341361621375666e-05, + "loss": 0.266, + "num_input_tokens_seen": 6406936, + "step": 23930 + }, + { + "epoch": 6.27391874180865, + "grad_norm": 1.095799207687378, + "learning_rate": 1.8330338178370706e-05, + "loss": 0.2305, + "num_input_tokens_seen": 6408216, + "step": 23935 + }, + { + "epoch": 6.275229357798165, + "grad_norm": 1.8445329666137695, + "learning_rate": 1.8319316131314447e-05, + "loss": 0.195, + "num_input_tokens_seen": 6409208, + "step": 23940 + }, + { + "epoch": 6.27653997378768, + "grad_norm": 1.0817383527755737, + "learning_rate": 1.830829548251377e-05, + "loss": 0.2469, + "num_input_tokens_seen": 6411704, + "step": 23945 + }, + { + "epoch": 6.277850589777195, + "grad_norm": 5.010278701782227, + "learning_rate": 1.829727623427529e-05, + "loss": 0.2713, + "num_input_tokens_seen": 6413256, + "step": 23950 + }, + { + "epoch": 6.2791612057667106, + "grad_norm": 0.934780478477478, + "learning_rate": 1.8286258388905308e-05, + "loss": 0.1597, + "num_input_tokens_seen": 6414440, + "step": 23955 + }, + { + "epoch": 6.280471821756225, + "grad_norm": 1.863503098487854, + "learning_rate": 1.8275241948709847e-05, + "loss": 0.2247, + "num_input_tokens_seen": 6415336, + "step": 23960 + }, + { + "epoch": 6.281782437745741, + "grad_norm": 1.6235357522964478, + "learning_rate": 1.8264226915994627e-05, + "loss": 0.1673, + "num_input_tokens_seen": 6417224, + "step": 23965 + }, + { + "epoch": 6.283093053735255, + "grad_norm": 1.682081937789917, + "learning_rate": 1.8253213293065077e-05, + "loss": 0.2593, + "num_input_tokens_seen": 6418456, + "step": 23970 + }, + { + "epoch": 6.284403669724771, + "grad_norm": 5.317943572998047, + "learning_rate": 1.8242201082226332e-05, + "loss": 0.3963, + "num_input_tokens_seen": 6419768, + "step": 23975 + }, + { + "epoch": 6.285714285714286, + "grad_norm": 1.7099953889846802, + "learning_rate": 1.8231190285783216e-05, + "loss": 0.4556, + "num_input_tokens_seen": 6421128, + "step": 23980 + }, + { + "epoch": 6.287024901703801, + "grad_norm": 1.8937121629714966, + "learning_rate": 1.822018090604028e-05, + "loss": 0.2993, + "num_input_tokens_seen": 6422712, + "step": 23985 + }, + { + "epoch": 6.288335517693316, + "grad_norm": 1.3581445217132568, + "learning_rate": 1.8209172945301768e-05, + "loss": 0.2144, + "num_input_tokens_seen": 6423608, + "step": 23990 + }, + { + "epoch": 6.289646133682831, + "grad_norm": 1.1810121536254883, + "learning_rate": 1.819816640587162e-05, + "loss": 0.205, + "num_input_tokens_seen": 6424888, + "step": 23995 + }, + { + "epoch": 6.290956749672346, + "grad_norm": 4.282624244689941, + "learning_rate": 1.8187161290053498e-05, + "loss": 0.2622, + "num_input_tokens_seen": 6425896, + "step": 24000 + }, + { + "epoch": 6.292267365661861, + "grad_norm": 1.2148356437683105, + "learning_rate": 1.817615760015074e-05, + "loss": 0.1905, + "num_input_tokens_seen": 6427144, + "step": 24005 + }, + { + "epoch": 6.293577981651376, + "grad_norm": 1.1339167356491089, + "learning_rate": 1.8165155338466404e-05, + "loss": 0.1547, + "num_input_tokens_seen": 6428424, + "step": 24010 + }, + { + "epoch": 6.294888597640891, + "grad_norm": 2.2519683837890625, + "learning_rate": 1.8154154507303245e-05, + "loss": 0.1584, + "num_input_tokens_seen": 6429848, + "step": 24015 + }, + { + "epoch": 6.296199213630406, + "grad_norm": 2.4602575302124023, + "learning_rate": 1.814315510896372e-05, + "loss": 0.2728, + "num_input_tokens_seen": 6431096, + "step": 24020 + }, + { + "epoch": 6.297509829619921, + "grad_norm": 2.056795358657837, + "learning_rate": 1.813215714574999e-05, + "loss": 0.2323, + "num_input_tokens_seen": 6432232, + "step": 24025 + }, + { + "epoch": 6.2988204456094365, + "grad_norm": 1.9023224115371704, + "learning_rate": 1.8121160619963894e-05, + "loss": 0.1682, + "num_input_tokens_seen": 6433624, + "step": 24030 + }, + { + "epoch": 6.300131061598951, + "grad_norm": 1.296014428138733, + "learning_rate": 1.8110165533907007e-05, + "loss": 0.2476, + "num_input_tokens_seen": 6434696, + "step": 24035 + }, + { + "epoch": 6.301441677588467, + "grad_norm": 1.1686640977859497, + "learning_rate": 1.809917188988056e-05, + "loss": 0.2018, + "num_input_tokens_seen": 6436104, + "step": 24040 + }, + { + "epoch": 6.302752293577981, + "grad_norm": 2.4388229846954346, + "learning_rate": 1.8088179690185515e-05, + "loss": 0.2909, + "num_input_tokens_seen": 6437336, + "step": 24045 + }, + { + "epoch": 6.304062909567497, + "grad_norm": 1.5905979871749878, + "learning_rate": 1.8077188937122515e-05, + "loss": 0.1841, + "num_input_tokens_seen": 6438504, + "step": 24050 + }, + { + "epoch": 6.305373525557012, + "grad_norm": 0.8201033473014832, + "learning_rate": 1.8066199632991913e-05, + "loss": 0.2193, + "num_input_tokens_seen": 6439592, + "step": 24055 + }, + { + "epoch": 6.306684141546527, + "grad_norm": 1.271535038948059, + "learning_rate": 1.805521178009375e-05, + "loss": 0.3532, + "num_input_tokens_seen": 6441000, + "step": 24060 + }, + { + "epoch": 6.307994757536042, + "grad_norm": 1.5571742057800293, + "learning_rate": 1.804422538072775e-05, + "loss": 0.2159, + "num_input_tokens_seen": 6442376, + "step": 24065 + }, + { + "epoch": 6.309305373525557, + "grad_norm": 1.7660539150238037, + "learning_rate": 1.8033240437193364e-05, + "loss": 0.1965, + "num_input_tokens_seen": 6443688, + "step": 24070 + }, + { + "epoch": 6.310615989515072, + "grad_norm": 1.55291748046875, + "learning_rate": 1.8022256951789718e-05, + "loss": 0.2984, + "num_input_tokens_seen": 6445176, + "step": 24075 + }, + { + "epoch": 6.3119266055045875, + "grad_norm": 1.3007245063781738, + "learning_rate": 1.8011274926815635e-05, + "loss": 0.1791, + "num_input_tokens_seen": 6446360, + "step": 24080 + }, + { + "epoch": 6.313237221494102, + "grad_norm": 3.2833220958709717, + "learning_rate": 1.8000294364569636e-05, + "loss": 0.487, + "num_input_tokens_seen": 6447560, + "step": 24085 + }, + { + "epoch": 6.314547837483618, + "grad_norm": 1.2997260093688965, + "learning_rate": 1.7989315267349936e-05, + "loss": 0.2176, + "num_input_tokens_seen": 6448792, + "step": 24090 + }, + { + "epoch": 6.315858453473132, + "grad_norm": 1.1953227519989014, + "learning_rate": 1.7978337637454433e-05, + "loss": 0.1891, + "num_input_tokens_seen": 6450232, + "step": 24095 + }, + { + "epoch": 6.317169069462647, + "grad_norm": 1.1176650524139404, + "learning_rate": 1.7967361477180727e-05, + "loss": 0.2986, + "num_input_tokens_seen": 6451736, + "step": 24100 + }, + { + "epoch": 6.3184796854521625, + "grad_norm": 2.2503249645233154, + "learning_rate": 1.7956386788826112e-05, + "loss": 0.2703, + "num_input_tokens_seen": 6452696, + "step": 24105 + }, + { + "epoch": 6.319790301441677, + "grad_norm": 1.6151015758514404, + "learning_rate": 1.7945413574687574e-05, + "loss": 0.3036, + "num_input_tokens_seen": 6454168, + "step": 24110 + }, + { + "epoch": 6.321100917431193, + "grad_norm": 3.063737630844116, + "learning_rate": 1.7934441837061795e-05, + "loss": 0.3831, + "num_input_tokens_seen": 6455480, + "step": 24115 + }, + { + "epoch": 6.322411533420707, + "grad_norm": 2.568227529525757, + "learning_rate": 1.792347157824512e-05, + "loss": 0.1871, + "num_input_tokens_seen": 6456728, + "step": 24120 + }, + { + "epoch": 6.323722149410223, + "grad_norm": 2.15976881980896, + "learning_rate": 1.7912502800533604e-05, + "loss": 0.2614, + "num_input_tokens_seen": 6458152, + "step": 24125 + }, + { + "epoch": 6.325032765399738, + "grad_norm": 1.8709485530853271, + "learning_rate": 1.7901535506223027e-05, + "loss": 0.2017, + "num_input_tokens_seen": 6459304, + "step": 24130 + }, + { + "epoch": 6.326343381389253, + "grad_norm": 1.5192255973815918, + "learning_rate": 1.7890569697608793e-05, + "loss": 0.3047, + "num_input_tokens_seen": 6460376, + "step": 24135 + }, + { + "epoch": 6.327653997378768, + "grad_norm": 1.6850937604904175, + "learning_rate": 1.787960537698604e-05, + "loss": 0.2369, + "num_input_tokens_seen": 6461528, + "step": 24140 + }, + { + "epoch": 6.328964613368283, + "grad_norm": 1.3149982690811157, + "learning_rate": 1.786864254664958e-05, + "loss": 0.1855, + "num_input_tokens_seen": 6463512, + "step": 24145 + }, + { + "epoch": 6.330275229357798, + "grad_norm": 1.0990643501281738, + "learning_rate": 1.785768120889392e-05, + "loss": 0.1832, + "num_input_tokens_seen": 6464568, + "step": 24150 + }, + { + "epoch": 6.3315858453473135, + "grad_norm": 2.3189496994018555, + "learning_rate": 1.784672136601323e-05, + "loss": 0.252, + "num_input_tokens_seen": 6465688, + "step": 24155 + }, + { + "epoch": 6.332896461336828, + "grad_norm": 0.7709469795227051, + "learning_rate": 1.7835763020301403e-05, + "loss": 0.1645, + "num_input_tokens_seen": 6467288, + "step": 24160 + }, + { + "epoch": 6.334207077326344, + "grad_norm": 2.858640432357788, + "learning_rate": 1.7824806174051995e-05, + "loss": 0.1554, + "num_input_tokens_seen": 6468232, + "step": 24165 + }, + { + "epoch": 6.335517693315858, + "grad_norm": 3.6545183658599854, + "learning_rate": 1.7813850829558257e-05, + "loss": 0.2408, + "num_input_tokens_seen": 6469512, + "step": 24170 + }, + { + "epoch": 6.336828309305374, + "grad_norm": 5.675177097320557, + "learning_rate": 1.7802896989113127e-05, + "loss": 0.2936, + "num_input_tokens_seen": 6470744, + "step": 24175 + }, + { + "epoch": 6.3381389252948885, + "grad_norm": 2.3556764125823975, + "learning_rate": 1.779194465500922e-05, + "loss": 0.2489, + "num_input_tokens_seen": 6471864, + "step": 24180 + }, + { + "epoch": 6.339449541284404, + "grad_norm": 3.8835318088531494, + "learning_rate": 1.7780993829538836e-05, + "loss": 0.2717, + "num_input_tokens_seen": 6472904, + "step": 24185 + }, + { + "epoch": 6.340760157273919, + "grad_norm": 1.3461283445358276, + "learning_rate": 1.7770044514993973e-05, + "loss": 0.2948, + "num_input_tokens_seen": 6474808, + "step": 24190 + }, + { + "epoch": 6.342070773263433, + "grad_norm": 1.92019784450531, + "learning_rate": 1.7759096713666305e-05, + "loss": 0.4005, + "num_input_tokens_seen": 6476168, + "step": 24195 + }, + { + "epoch": 6.343381389252949, + "grad_norm": 0.6912484765052795, + "learning_rate": 1.774815042784718e-05, + "loss": 0.1159, + "num_input_tokens_seen": 6477160, + "step": 24200 + }, + { + "epoch": 6.344692005242464, + "grad_norm": 3.0113747119903564, + "learning_rate": 1.773720565982764e-05, + "loss": 0.2619, + "num_input_tokens_seen": 6478600, + "step": 24205 + }, + { + "epoch": 6.346002621231979, + "grad_norm": 1.4642149209976196, + "learning_rate": 1.7726262411898413e-05, + "loss": 0.2731, + "num_input_tokens_seen": 6480408, + "step": 24210 + }, + { + "epoch": 6.347313237221494, + "grad_norm": 4.554865837097168, + "learning_rate": 1.771532068634989e-05, + "loss": 0.3357, + "num_input_tokens_seen": 6481416, + "step": 24215 + }, + { + "epoch": 6.348623853211009, + "grad_norm": 2.037632465362549, + "learning_rate": 1.770438048547216e-05, + "loss": 0.4477, + "num_input_tokens_seen": 6482840, + "step": 24220 + }, + { + "epoch": 6.349934469200524, + "grad_norm": 1.6537226438522339, + "learning_rate": 1.769344181155499e-05, + "loss": 0.294, + "num_input_tokens_seen": 6484024, + "step": 24225 + }, + { + "epoch": 6.3512450851900395, + "grad_norm": 1.7044411897659302, + "learning_rate": 1.7682504666887823e-05, + "loss": 0.2279, + "num_input_tokens_seen": 6485192, + "step": 24230 + }, + { + "epoch": 6.352555701179554, + "grad_norm": 1.3577054738998413, + "learning_rate": 1.767156905375979e-05, + "loss": 0.1947, + "num_input_tokens_seen": 6486760, + "step": 24235 + }, + { + "epoch": 6.35386631716907, + "grad_norm": 2.3736534118652344, + "learning_rate": 1.7660634974459674e-05, + "loss": 0.2758, + "num_input_tokens_seen": 6487960, + "step": 24240 + }, + { + "epoch": 6.355176933158584, + "grad_norm": 1.8477109670639038, + "learning_rate": 1.7649702431275994e-05, + "loss": 0.4716, + "num_input_tokens_seen": 6489096, + "step": 24245 + }, + { + "epoch": 6.3564875491481, + "grad_norm": 1.122439980506897, + "learning_rate": 1.7638771426496893e-05, + "loss": 0.2635, + "num_input_tokens_seen": 6490344, + "step": 24250 + }, + { + "epoch": 6.3577981651376145, + "grad_norm": 2.35628604888916, + "learning_rate": 1.7627841962410208e-05, + "loss": 0.3211, + "num_input_tokens_seen": 6491832, + "step": 24255 + }, + { + "epoch": 6.35910878112713, + "grad_norm": 9.740086555480957, + "learning_rate": 1.7616914041303463e-05, + "loss": 0.324, + "num_input_tokens_seen": 6492952, + "step": 24260 + }, + { + "epoch": 6.360419397116645, + "grad_norm": 2.354307174682617, + "learning_rate": 1.7605987665463854e-05, + "loss": 0.1863, + "num_input_tokens_seen": 6494248, + "step": 24265 + }, + { + "epoch": 6.36173001310616, + "grad_norm": 0.5764444470405579, + "learning_rate": 1.7595062837178254e-05, + "loss": 0.131, + "num_input_tokens_seen": 6495336, + "step": 24270 + }, + { + "epoch": 6.363040629095675, + "grad_norm": 1.5795695781707764, + "learning_rate": 1.7584139558733205e-05, + "loss": 0.1565, + "num_input_tokens_seen": 6496312, + "step": 24275 + }, + { + "epoch": 6.3643512450851905, + "grad_norm": 1.4334341287612915, + "learning_rate": 1.757321783241493e-05, + "loss": 0.2096, + "num_input_tokens_seen": 6497528, + "step": 24280 + }, + { + "epoch": 6.365661861074705, + "grad_norm": 1.0731923580169678, + "learning_rate": 1.7562297660509334e-05, + "loss": 0.3641, + "num_input_tokens_seen": 6498936, + "step": 24285 + }, + { + "epoch": 6.36697247706422, + "grad_norm": 1.4045175313949585, + "learning_rate": 1.7551379045301987e-05, + "loss": 0.2273, + "num_input_tokens_seen": 6500024, + "step": 24290 + }, + { + "epoch": 6.368283093053735, + "grad_norm": 1.1564463376998901, + "learning_rate": 1.7540461989078132e-05, + "loss": 0.3539, + "num_input_tokens_seen": 6501640, + "step": 24295 + }, + { + "epoch": 6.36959370904325, + "grad_norm": 1.1792147159576416, + "learning_rate": 1.7529546494122685e-05, + "loss": 0.2426, + "num_input_tokens_seen": 6503144, + "step": 24300 + }, + { + "epoch": 6.3709043250327655, + "grad_norm": 1.8816765546798706, + "learning_rate": 1.7518632562720263e-05, + "loss": 0.3061, + "num_input_tokens_seen": 6504360, + "step": 24305 + }, + { + "epoch": 6.37221494102228, + "grad_norm": 1.1695934534072876, + "learning_rate": 1.7507720197155114e-05, + "loss": 0.2366, + "num_input_tokens_seen": 6505992, + "step": 24310 + }, + { + "epoch": 6.373525557011796, + "grad_norm": 1.1235312223434448, + "learning_rate": 1.749680939971118e-05, + "loss": 0.4744, + "num_input_tokens_seen": 6507608, + "step": 24315 + }, + { + "epoch": 6.37483617300131, + "grad_norm": 2.192948579788208, + "learning_rate": 1.7485900172672076e-05, + "loss": 0.2382, + "num_input_tokens_seen": 6509112, + "step": 24320 + }, + { + "epoch": 6.376146788990826, + "grad_norm": 1.1891158819198608, + "learning_rate": 1.7474992518321083e-05, + "loss": 0.1497, + "num_input_tokens_seen": 6510408, + "step": 24325 + }, + { + "epoch": 6.3774574049803405, + "grad_norm": 3.32666277885437, + "learning_rate": 1.7464086438941156e-05, + "loss": 0.417, + "num_input_tokens_seen": 6511992, + "step": 24330 + }, + { + "epoch": 6.378768020969856, + "grad_norm": 1.9971271753311157, + "learning_rate": 1.7453181936814915e-05, + "loss": 0.2196, + "num_input_tokens_seen": 6513288, + "step": 24335 + }, + { + "epoch": 6.380078636959371, + "grad_norm": 2.5925989151000977, + "learning_rate": 1.7442279014224654e-05, + "loss": 0.2782, + "num_input_tokens_seen": 6514456, + "step": 24340 + }, + { + "epoch": 6.381389252948886, + "grad_norm": 2.790924310684204, + "learning_rate": 1.7431377673452336e-05, + "loss": 0.2153, + "num_input_tokens_seen": 6515832, + "step": 24345 + }, + { + "epoch": 6.382699868938401, + "grad_norm": 0.9551293849945068, + "learning_rate": 1.74204779167796e-05, + "loss": 0.2869, + "num_input_tokens_seen": 6517000, + "step": 24350 + }, + { + "epoch": 6.3840104849279165, + "grad_norm": 1.9330735206604004, + "learning_rate": 1.740957974648772e-05, + "loss": 0.2947, + "num_input_tokens_seen": 6519128, + "step": 24355 + }, + { + "epoch": 6.385321100917431, + "grad_norm": 0.9828119277954102, + "learning_rate": 1.7398683164857697e-05, + "loss": 0.1577, + "num_input_tokens_seen": 6520104, + "step": 24360 + }, + { + "epoch": 6.386631716906947, + "grad_norm": 1.5239214897155762, + "learning_rate": 1.7387788174170157e-05, + "loss": 0.2763, + "num_input_tokens_seen": 6521720, + "step": 24365 + }, + { + "epoch": 6.387942332896461, + "grad_norm": 1.2978180646896362, + "learning_rate": 1.7376894776705397e-05, + "loss": 0.2115, + "num_input_tokens_seen": 6523304, + "step": 24370 + }, + { + "epoch": 6.389252948885977, + "grad_norm": 1.5011001825332642, + "learning_rate": 1.7366002974743383e-05, + "loss": 0.2122, + "num_input_tokens_seen": 6524536, + "step": 24375 + }, + { + "epoch": 6.3905635648754915, + "grad_norm": 1.3320910930633545, + "learning_rate": 1.7355112770563764e-05, + "loss": 0.3208, + "num_input_tokens_seen": 6526024, + "step": 24380 + }, + { + "epoch": 6.391874180865006, + "grad_norm": 4.077765941619873, + "learning_rate": 1.734422416644583e-05, + "loss": 0.3404, + "num_input_tokens_seen": 6527112, + "step": 24385 + }, + { + "epoch": 6.393184796854522, + "grad_norm": 1.944440484046936, + "learning_rate": 1.733333716466855e-05, + "loss": 0.34, + "num_input_tokens_seen": 6528552, + "step": 24390 + }, + { + "epoch": 6.394495412844036, + "grad_norm": 4.191859722137451, + "learning_rate": 1.732245176751055e-05, + "loss": 0.2912, + "num_input_tokens_seen": 6529688, + "step": 24395 + }, + { + "epoch": 6.395806028833552, + "grad_norm": 2.406080484390259, + "learning_rate": 1.7311567977250133e-05, + "loss": 0.3077, + "num_input_tokens_seen": 6531032, + "step": 24400 + }, + { + "epoch": 6.3971166448230665, + "grad_norm": 3.5252606868743896, + "learning_rate": 1.7300685796165255e-05, + "loss": 0.1641, + "num_input_tokens_seen": 6531944, + "step": 24405 + }, + { + "epoch": 6.398427260812582, + "grad_norm": 0.38855546712875366, + "learning_rate": 1.728980522653354e-05, + "loss": 0.1929, + "num_input_tokens_seen": 6532968, + "step": 24410 + }, + { + "epoch": 6.399737876802097, + "grad_norm": 2.313277244567871, + "learning_rate": 1.727892627063225e-05, + "loss": 0.2341, + "num_input_tokens_seen": 6534152, + "step": 24415 + }, + { + "epoch": 6.401048492791612, + "grad_norm": 1.358633279800415, + "learning_rate": 1.726804893073837e-05, + "loss": 0.2893, + "num_input_tokens_seen": 6535608, + "step": 24420 + }, + { + "epoch": 6.402359108781127, + "grad_norm": 1.501181960105896, + "learning_rate": 1.725717320912848e-05, + "loss": 0.3891, + "num_input_tokens_seen": 6536760, + "step": 24425 + }, + { + "epoch": 6.4036697247706424, + "grad_norm": 2.2176167964935303, + "learning_rate": 1.724629910807886e-05, + "loss": 0.3451, + "num_input_tokens_seen": 6538232, + "step": 24430 + }, + { + "epoch": 6.404980340760157, + "grad_norm": 4.613729000091553, + "learning_rate": 1.7235426629865445e-05, + "loss": 0.1314, + "num_input_tokens_seen": 6539464, + "step": 24435 + }, + { + "epoch": 6.406290956749673, + "grad_norm": 1.439953088760376, + "learning_rate": 1.7224555776763814e-05, + "loss": 0.191, + "num_input_tokens_seen": 6540568, + "step": 24440 + }, + { + "epoch": 6.407601572739187, + "grad_norm": 2.7714107036590576, + "learning_rate": 1.7213686551049233e-05, + "loss": 0.2988, + "num_input_tokens_seen": 6543944, + "step": 24445 + }, + { + "epoch": 6.408912188728703, + "grad_norm": 1.582044243812561, + "learning_rate": 1.7202818954996597e-05, + "loss": 0.2112, + "num_input_tokens_seen": 6544920, + "step": 24450 + }, + { + "epoch": 6.4102228047182175, + "grad_norm": 2.488283157348633, + "learning_rate": 1.7191952990880476e-05, + "loss": 0.265, + "num_input_tokens_seen": 6546136, + "step": 24455 + }, + { + "epoch": 6.411533420707733, + "grad_norm": 2.0571134090423584, + "learning_rate": 1.7181088660975107e-05, + "loss": 0.24, + "num_input_tokens_seen": 6547272, + "step": 24460 + }, + { + "epoch": 6.412844036697248, + "grad_norm": 1.4961063861846924, + "learning_rate": 1.7170225967554374e-05, + "loss": 0.176, + "num_input_tokens_seen": 6548376, + "step": 24465 + }, + { + "epoch": 6.414154652686763, + "grad_norm": 1.1828079223632812, + "learning_rate": 1.715936491289181e-05, + "loss": 0.2579, + "num_input_tokens_seen": 6550072, + "step": 24470 + }, + { + "epoch": 6.415465268676278, + "grad_norm": 1.1920950412750244, + "learning_rate": 1.7148505499260627e-05, + "loss": 0.2947, + "num_input_tokens_seen": 6551736, + "step": 24475 + }, + { + "epoch": 6.4167758846657925, + "grad_norm": 0.8958536386489868, + "learning_rate": 1.713764772893368e-05, + "loss": 0.242, + "num_input_tokens_seen": 6553144, + "step": 24480 + }, + { + "epoch": 6.418086500655308, + "grad_norm": 2.8996334075927734, + "learning_rate": 1.7126791604183474e-05, + "loss": 0.2276, + "num_input_tokens_seen": 6554152, + "step": 24485 + }, + { + "epoch": 6.419397116644823, + "grad_norm": 1.4183218479156494, + "learning_rate": 1.7115937127282182e-05, + "loss": 0.1903, + "num_input_tokens_seen": 6555160, + "step": 24490 + }, + { + "epoch": 6.420707732634338, + "grad_norm": 1.6993072032928467, + "learning_rate": 1.7105084300501627e-05, + "loss": 0.2461, + "num_input_tokens_seen": 6556232, + "step": 24495 + }, + { + "epoch": 6.422018348623853, + "grad_norm": 1.4693557024002075, + "learning_rate": 1.7094233126113285e-05, + "loss": 0.3941, + "num_input_tokens_seen": 6557448, + "step": 24500 + }, + { + "epoch": 6.423328964613368, + "grad_norm": 8.028271675109863, + "learning_rate": 1.70833836063883e-05, + "loss": 0.2356, + "num_input_tokens_seen": 6558680, + "step": 24505 + }, + { + "epoch": 6.424639580602883, + "grad_norm": 2.5890910625457764, + "learning_rate": 1.7072535743597433e-05, + "loss": 0.3006, + "num_input_tokens_seen": 6559848, + "step": 24510 + }, + { + "epoch": 6.425950196592399, + "grad_norm": 1.4581122398376465, + "learning_rate": 1.7061689540011144e-05, + "loss": 0.1915, + "num_input_tokens_seen": 6560904, + "step": 24515 + }, + { + "epoch": 6.427260812581913, + "grad_norm": 2.638091802597046, + "learning_rate": 1.7050844997899517e-05, + "loss": 0.2399, + "num_input_tokens_seen": 6562184, + "step": 24520 + }, + { + "epoch": 6.428571428571429, + "grad_norm": 2.012669324874878, + "learning_rate": 1.70400021195323e-05, + "loss": 0.2031, + "num_input_tokens_seen": 6563656, + "step": 24525 + }, + { + "epoch": 6.4298820445609435, + "grad_norm": 2.179912805557251, + "learning_rate": 1.702916090717887e-05, + "loss": 0.2655, + "num_input_tokens_seen": 6564808, + "step": 24530 + }, + { + "epoch": 6.431192660550459, + "grad_norm": 1.7181193828582764, + "learning_rate": 1.7018321363108303e-05, + "loss": 0.2304, + "num_input_tokens_seen": 6565656, + "step": 24535 + }, + { + "epoch": 6.432503276539974, + "grad_norm": 1.4956001043319702, + "learning_rate": 1.700748348958929e-05, + "loss": 0.2822, + "num_input_tokens_seen": 6567176, + "step": 24540 + }, + { + "epoch": 6.433813892529489, + "grad_norm": 2.421895742416382, + "learning_rate": 1.699664728889016e-05, + "loss": 0.2212, + "num_input_tokens_seen": 6568520, + "step": 24545 + }, + { + "epoch": 6.435124508519004, + "grad_norm": 2.019301414489746, + "learning_rate": 1.698581276327893e-05, + "loss": 0.2069, + "num_input_tokens_seen": 6569592, + "step": 24550 + }, + { + "epoch": 6.436435124508519, + "grad_norm": 2.262328863143921, + "learning_rate": 1.6974979915023237e-05, + "loss": 0.2036, + "num_input_tokens_seen": 6570744, + "step": 24555 + }, + { + "epoch": 6.437745740498034, + "grad_norm": 2.1716225147247314, + "learning_rate": 1.6964148746390383e-05, + "loss": 0.1597, + "num_input_tokens_seen": 6571704, + "step": 24560 + }, + { + "epoch": 6.43905635648755, + "grad_norm": 1.5705296993255615, + "learning_rate": 1.6953319259647323e-05, + "loss": 0.3337, + "num_input_tokens_seen": 6572904, + "step": 24565 + }, + { + "epoch": 6.440366972477064, + "grad_norm": 2.115091323852539, + "learning_rate": 1.6942491457060626e-05, + "loss": 0.4859, + "num_input_tokens_seen": 6574520, + "step": 24570 + }, + { + "epoch": 6.441677588466579, + "grad_norm": 2.8258392810821533, + "learning_rate": 1.6931665340896545e-05, + "loss": 0.2933, + "num_input_tokens_seen": 6576152, + "step": 24575 + }, + { + "epoch": 6.442988204456094, + "grad_norm": 1.5833500623703003, + "learning_rate": 1.692084091342097e-05, + "loss": 0.2903, + "num_input_tokens_seen": 6577512, + "step": 24580 + }, + { + "epoch": 6.444298820445609, + "grad_norm": 0.5947524905204773, + "learning_rate": 1.6910018176899424e-05, + "loss": 0.1503, + "num_input_tokens_seen": 6578648, + "step": 24585 + }, + { + "epoch": 6.445609436435125, + "grad_norm": 1.279109239578247, + "learning_rate": 1.68991971335971e-05, + "loss": 0.2222, + "num_input_tokens_seen": 6579992, + "step": 24590 + }, + { + "epoch": 6.446920052424639, + "grad_norm": 1.6015174388885498, + "learning_rate": 1.688837778577882e-05, + "loss": 0.1797, + "num_input_tokens_seen": 6581096, + "step": 24595 + }, + { + "epoch": 6.448230668414155, + "grad_norm": 4.362173557281494, + "learning_rate": 1.687756013570906e-05, + "loss": 0.2566, + "num_input_tokens_seen": 6582344, + "step": 24600 + }, + { + "epoch": 6.4495412844036695, + "grad_norm": 2.9380645751953125, + "learning_rate": 1.6866744185651922e-05, + "loss": 0.2607, + "num_input_tokens_seen": 6583800, + "step": 24605 + }, + { + "epoch": 6.450851900393185, + "grad_norm": 1.0816841125488281, + "learning_rate": 1.6855929937871175e-05, + "loss": 0.2794, + "num_input_tokens_seen": 6585896, + "step": 24610 + }, + { + "epoch": 6.4521625163827, + "grad_norm": 1.7852445840835571, + "learning_rate": 1.6845117394630223e-05, + "loss": 0.2299, + "num_input_tokens_seen": 6587224, + "step": 24615 + }, + { + "epoch": 6.453473132372215, + "grad_norm": 1.5401229858398438, + "learning_rate": 1.683430655819211e-05, + "loss": 0.274, + "num_input_tokens_seen": 6588792, + "step": 24620 + }, + { + "epoch": 6.45478374836173, + "grad_norm": 1.7016922235488892, + "learning_rate": 1.6823497430819518e-05, + "loss": 0.2232, + "num_input_tokens_seen": 6590104, + "step": 24625 + }, + { + "epoch": 6.456094364351245, + "grad_norm": 2.0702288150787354, + "learning_rate": 1.681269001477479e-05, + "loss": 0.2478, + "num_input_tokens_seen": 6591032, + "step": 24630 + }, + { + "epoch": 6.45740498034076, + "grad_norm": 1.1269571781158447, + "learning_rate": 1.6801884312319895e-05, + "loss": 0.2801, + "num_input_tokens_seen": 6592664, + "step": 24635 + }, + { + "epoch": 6.458715596330276, + "grad_norm": 0.8504106402397156, + "learning_rate": 1.6791080325716445e-05, + "loss": 0.2147, + "num_input_tokens_seen": 6594184, + "step": 24640 + }, + { + "epoch": 6.46002621231979, + "grad_norm": 1.611714243888855, + "learning_rate": 1.6780278057225697e-05, + "loss": 0.1609, + "num_input_tokens_seen": 6595320, + "step": 24645 + }, + { + "epoch": 6.461336828309306, + "grad_norm": 1.5134468078613281, + "learning_rate": 1.6769477509108556e-05, + "loss": 0.3378, + "num_input_tokens_seen": 6596680, + "step": 24650 + }, + { + "epoch": 6.46264744429882, + "grad_norm": 1.8676031827926636, + "learning_rate": 1.675867868362555e-05, + "loss": 0.2921, + "num_input_tokens_seen": 6598840, + "step": 24655 + }, + { + "epoch": 6.463958060288336, + "grad_norm": 1.5996061563491821, + "learning_rate": 1.6747881583036858e-05, + "loss": 0.2671, + "num_input_tokens_seen": 6600088, + "step": 24660 + }, + { + "epoch": 6.465268676277851, + "grad_norm": 4.331786632537842, + "learning_rate": 1.673708620960229e-05, + "loss": 0.2526, + "num_input_tokens_seen": 6601320, + "step": 24665 + }, + { + "epoch": 6.466579292267365, + "grad_norm": 3.348104238510132, + "learning_rate": 1.6726292565581298e-05, + "loss": 0.1584, + "num_input_tokens_seen": 6602376, + "step": 24670 + }, + { + "epoch": 6.467889908256881, + "grad_norm": 3.22426438331604, + "learning_rate": 1.6715500653232974e-05, + "loss": 0.2319, + "num_input_tokens_seen": 6604344, + "step": 24675 + }, + { + "epoch": 6.4692005242463955, + "grad_norm": 1.2351508140563965, + "learning_rate": 1.670471047481606e-05, + "loss": 0.2927, + "num_input_tokens_seen": 6606088, + "step": 24680 + }, + { + "epoch": 6.470511140235911, + "grad_norm": 1.6564446687698364, + "learning_rate": 1.66939220325889e-05, + "loss": 0.1923, + "num_input_tokens_seen": 6607560, + "step": 24685 + }, + { + "epoch": 6.471821756225426, + "grad_norm": 2.0077171325683594, + "learning_rate": 1.6683135328809512e-05, + "loss": 0.2636, + "num_input_tokens_seen": 6608904, + "step": 24690 + }, + { + "epoch": 6.473132372214941, + "grad_norm": 1.3629307746887207, + "learning_rate": 1.6672350365735528e-05, + "loss": 0.2658, + "num_input_tokens_seen": 6610088, + "step": 24695 + }, + { + "epoch": 6.474442988204456, + "grad_norm": 1.450796127319336, + "learning_rate": 1.6661567145624215e-05, + "loss": 0.3331, + "num_input_tokens_seen": 6611640, + "step": 24700 + }, + { + "epoch": 6.475753604193971, + "grad_norm": 2.8802084922790527, + "learning_rate": 1.6650785670732498e-05, + "loss": 0.242, + "num_input_tokens_seen": 6612968, + "step": 24705 + }, + { + "epoch": 6.477064220183486, + "grad_norm": 2.166536808013916, + "learning_rate": 1.6640005943316914e-05, + "loss": 0.1642, + "num_input_tokens_seen": 6614072, + "step": 24710 + }, + { + "epoch": 6.478374836173002, + "grad_norm": 1.056753396987915, + "learning_rate": 1.6629227965633644e-05, + "loss": 0.1821, + "num_input_tokens_seen": 6615384, + "step": 24715 + }, + { + "epoch": 6.479685452162516, + "grad_norm": 2.0155320167541504, + "learning_rate": 1.6618451739938495e-05, + "loss": 0.3327, + "num_input_tokens_seen": 6616904, + "step": 24720 + }, + { + "epoch": 6.480996068152032, + "grad_norm": 3.5713560581207275, + "learning_rate": 1.6607677268486914e-05, + "loss": 0.5217, + "num_input_tokens_seen": 6618152, + "step": 24725 + }, + { + "epoch": 6.482306684141546, + "grad_norm": 1.8478790521621704, + "learning_rate": 1.659690455353398e-05, + "loss": 0.2993, + "num_input_tokens_seen": 6619432, + "step": 24730 + }, + { + "epoch": 6.483617300131062, + "grad_norm": 2.600252151489258, + "learning_rate": 1.65861335973344e-05, + "loss": 0.2132, + "num_input_tokens_seen": 6620760, + "step": 24735 + }, + { + "epoch": 6.484927916120577, + "grad_norm": 2.594769239425659, + "learning_rate": 1.6575364402142528e-05, + "loss": 0.3772, + "num_input_tokens_seen": 6622104, + "step": 24740 + }, + { + "epoch": 6.486238532110092, + "grad_norm": 0.6772327423095703, + "learning_rate": 1.6564596970212327e-05, + "loss": 0.2464, + "num_input_tokens_seen": 6623224, + "step": 24745 + }, + { + "epoch": 6.487549148099607, + "grad_norm": 1.1343536376953125, + "learning_rate": 1.6553831303797405e-05, + "loss": 0.2232, + "num_input_tokens_seen": 6624264, + "step": 24750 + }, + { + "epoch": 6.488859764089122, + "grad_norm": 2.823751211166382, + "learning_rate": 1.6543067405150994e-05, + "loss": 0.3139, + "num_input_tokens_seen": 6625240, + "step": 24755 + }, + { + "epoch": 6.490170380078637, + "grad_norm": 1.8242095708847046, + "learning_rate": 1.653230527652596e-05, + "loss": 0.2612, + "num_input_tokens_seen": 6626552, + "step": 24760 + }, + { + "epoch": 6.491480996068152, + "grad_norm": 1.2611595392227173, + "learning_rate": 1.6521544920174803e-05, + "loss": 0.3552, + "num_input_tokens_seen": 6627768, + "step": 24765 + }, + { + "epoch": 6.492791612057667, + "grad_norm": 0.7749793529510498, + "learning_rate": 1.6510786338349645e-05, + "loss": 0.2168, + "num_input_tokens_seen": 6629288, + "step": 24770 + }, + { + "epoch": 6.494102228047182, + "grad_norm": 1.5534158945083618, + "learning_rate": 1.6500029533302247e-05, + "loss": 0.3832, + "num_input_tokens_seen": 6630296, + "step": 24775 + }, + { + "epoch": 6.495412844036697, + "grad_norm": 1.4305404424667358, + "learning_rate": 1.648927450728397e-05, + "loss": 0.3007, + "num_input_tokens_seen": 6631400, + "step": 24780 + }, + { + "epoch": 6.496723460026212, + "grad_norm": 2.164033889770508, + "learning_rate": 1.6478521262545837e-05, + "loss": 0.2326, + "num_input_tokens_seen": 6632424, + "step": 24785 + }, + { + "epoch": 6.498034076015728, + "grad_norm": 1.021644115447998, + "learning_rate": 1.6467769801338478e-05, + "loss": 0.1606, + "num_input_tokens_seen": 6634120, + "step": 24790 + }, + { + "epoch": 6.499344692005242, + "grad_norm": 3.5551834106445312, + "learning_rate": 1.645702012591216e-05, + "loss": 0.227, + "num_input_tokens_seen": 6635256, + "step": 24795 + }, + { + "epoch": 6.500655307994758, + "grad_norm": 3.124446153640747, + "learning_rate": 1.644627223851677e-05, + "loss": 0.2133, + "num_input_tokens_seen": 6636552, + "step": 24800 + }, + { + "epoch": 6.50170380078637, + "eval_loss": 0.637935996055603, + "eval_runtime": 16.849, + "eval_samples_per_second": 50.329, + "eval_steps_per_second": 25.165, + "num_input_tokens_seen": 6637864, + "step": 24804 + }, + { + "epoch": 6.501965923984272, + "grad_norm": 1.8473727703094482, + "learning_rate": 1.643552614140182e-05, + "loss": 0.2731, + "num_input_tokens_seen": 6638008, + "step": 24805 + }, + { + "epoch": 6.503276539973788, + "grad_norm": 2.1088826656341553, + "learning_rate": 1.6424781836816443e-05, + "loss": 0.2276, + "num_input_tokens_seen": 6639304, + "step": 24810 + }, + { + "epoch": 6.504587155963303, + "grad_norm": 1.3382377624511719, + "learning_rate": 1.641403932700941e-05, + "loss": 0.2024, + "num_input_tokens_seen": 6640776, + "step": 24815 + }, + { + "epoch": 6.505897771952818, + "grad_norm": 1.1745539903640747, + "learning_rate": 1.6403298614229117e-05, + "loss": 0.2193, + "num_input_tokens_seen": 6642024, + "step": 24820 + }, + { + "epoch": 6.507208387942333, + "grad_norm": 2.989055633544922, + "learning_rate": 1.639255970072357e-05, + "loss": 0.2761, + "num_input_tokens_seen": 6643288, + "step": 24825 + }, + { + "epoch": 6.508519003931848, + "grad_norm": 3.1724979877471924, + "learning_rate": 1.6381822588740404e-05, + "loss": 0.2533, + "num_input_tokens_seen": 6644232, + "step": 24830 + }, + { + "epoch": 6.509829619921363, + "grad_norm": 1.1785675287246704, + "learning_rate": 1.6371087280526888e-05, + "loss": 0.162, + "num_input_tokens_seen": 6645496, + "step": 24835 + }, + { + "epoch": 6.511140235910878, + "grad_norm": 1.792145848274231, + "learning_rate": 1.6360353778329883e-05, + "loss": 0.2406, + "num_input_tokens_seen": 6647128, + "step": 24840 + }, + { + "epoch": 6.512450851900393, + "grad_norm": 2.407560110092163, + "learning_rate": 1.6349622084395903e-05, + "loss": 0.2655, + "num_input_tokens_seen": 6648056, + "step": 24845 + }, + { + "epoch": 6.513761467889909, + "grad_norm": 1.7401480674743652, + "learning_rate": 1.6338892200971078e-05, + "loss": 0.3267, + "num_input_tokens_seen": 6649176, + "step": 24850 + }, + { + "epoch": 6.515072083879423, + "grad_norm": 0.880858302116394, + "learning_rate": 1.6328164130301157e-05, + "loss": 0.2351, + "num_input_tokens_seen": 6650536, + "step": 24855 + }, + { + "epoch": 6.516382699868938, + "grad_norm": 1.2594380378723145, + "learning_rate": 1.631743787463149e-05, + "loss": 0.2694, + "num_input_tokens_seen": 6652072, + "step": 24860 + }, + { + "epoch": 6.517693315858454, + "grad_norm": 0.9371957778930664, + "learning_rate": 1.630671343620708e-05, + "loss": 0.1397, + "num_input_tokens_seen": 6653240, + "step": 24865 + }, + { + "epoch": 6.519003931847968, + "grad_norm": 1.800719976425171, + "learning_rate": 1.6295990817272516e-05, + "loss": 0.2653, + "num_input_tokens_seen": 6654440, + "step": 24870 + }, + { + "epoch": 6.520314547837484, + "grad_norm": 1.764522671699524, + "learning_rate": 1.6285270020072044e-05, + "loss": 0.209, + "num_input_tokens_seen": 6655608, + "step": 24875 + }, + { + "epoch": 6.521625163826998, + "grad_norm": 0.9336071610450745, + "learning_rate": 1.6274551046849495e-05, + "loss": 0.1293, + "num_input_tokens_seen": 6656712, + "step": 24880 + }, + { + "epoch": 6.522935779816514, + "grad_norm": 2.1324498653411865, + "learning_rate": 1.6263833899848337e-05, + "loss": 0.2684, + "num_input_tokens_seen": 6658936, + "step": 24885 + }, + { + "epoch": 6.524246395806029, + "grad_norm": 2.4640307426452637, + "learning_rate": 1.625311858131165e-05, + "loss": 0.3079, + "num_input_tokens_seen": 6660328, + "step": 24890 + }, + { + "epoch": 6.525557011795544, + "grad_norm": 2.192350149154663, + "learning_rate": 1.6242405093482133e-05, + "loss": 0.2818, + "num_input_tokens_seen": 6661352, + "step": 24895 + }, + { + "epoch": 6.526867627785059, + "grad_norm": 0.9033593535423279, + "learning_rate": 1.6231693438602097e-05, + "loss": 0.3324, + "num_input_tokens_seen": 6662600, + "step": 24900 + }, + { + "epoch": 6.528178243774574, + "grad_norm": 0.8199595212936401, + "learning_rate": 1.6220983618913472e-05, + "loss": 0.2879, + "num_input_tokens_seen": 6663656, + "step": 24905 + }, + { + "epoch": 6.529488859764089, + "grad_norm": 3.2966651916503906, + "learning_rate": 1.6210275636657805e-05, + "loss": 0.2283, + "num_input_tokens_seen": 6664744, + "step": 24910 + }, + { + "epoch": 6.5307994757536045, + "grad_norm": 1.1988612413406372, + "learning_rate": 1.6199569494076265e-05, + "loss": 0.2481, + "num_input_tokens_seen": 6666104, + "step": 24915 + }, + { + "epoch": 6.532110091743119, + "grad_norm": 2.3064749240875244, + "learning_rate": 1.618886519340962e-05, + "loss": 0.2656, + "num_input_tokens_seen": 6667352, + "step": 24920 + }, + { + "epoch": 6.533420707732635, + "grad_norm": 1.53348708152771, + "learning_rate": 1.6178162736898264e-05, + "loss": 0.3319, + "num_input_tokens_seen": 6669080, + "step": 24925 + }, + { + "epoch": 6.534731323722149, + "grad_norm": 1.9540153741836548, + "learning_rate": 1.61674621267822e-05, + "loss": 0.3535, + "num_input_tokens_seen": 6670712, + "step": 24930 + }, + { + "epoch": 6.536041939711664, + "grad_norm": 1.4071115255355835, + "learning_rate": 1.6156763365301054e-05, + "loss": 0.2354, + "num_input_tokens_seen": 6671944, + "step": 24935 + }, + { + "epoch": 6.53735255570118, + "grad_norm": 2.087322473526001, + "learning_rate": 1.614606645469406e-05, + "loss": 0.2616, + "num_input_tokens_seen": 6673080, + "step": 24940 + }, + { + "epoch": 6.538663171690695, + "grad_norm": 1.6254206895828247, + "learning_rate": 1.6135371397200057e-05, + "loss": 0.2947, + "num_input_tokens_seen": 6674056, + "step": 24945 + }, + { + "epoch": 6.53997378768021, + "grad_norm": 1.8674806356430054, + "learning_rate": 1.6124678195057507e-05, + "loss": 0.2196, + "num_input_tokens_seen": 6675368, + "step": 24950 + }, + { + "epoch": 6.541284403669724, + "grad_norm": 4.337184906005859, + "learning_rate": 1.6113986850504475e-05, + "loss": 0.3879, + "num_input_tokens_seen": 6676696, + "step": 24955 + }, + { + "epoch": 6.54259501965924, + "grad_norm": 1.4293012619018555, + "learning_rate": 1.6103297365778636e-05, + "loss": 0.3112, + "num_input_tokens_seen": 6678136, + "step": 24960 + }, + { + "epoch": 6.543905635648755, + "grad_norm": 2.40297269821167, + "learning_rate": 1.609260974311729e-05, + "loss": 0.2855, + "num_input_tokens_seen": 6679656, + "step": 24965 + }, + { + "epoch": 6.54521625163827, + "grad_norm": 1.2008287906646729, + "learning_rate": 1.6081923984757332e-05, + "loss": 0.2186, + "num_input_tokens_seen": 6680856, + "step": 24970 + }, + { + "epoch": 6.546526867627785, + "grad_norm": 2.6664834022521973, + "learning_rate": 1.607124009293528e-05, + "loss": 0.2262, + "num_input_tokens_seen": 6682296, + "step": 24975 + }, + { + "epoch": 6.5478374836173, + "grad_norm": 1.629953145980835, + "learning_rate": 1.6060558069887244e-05, + "loss": 0.223, + "num_input_tokens_seen": 6683624, + "step": 24980 + }, + { + "epoch": 6.549148099606815, + "grad_norm": 2.2630648612976074, + "learning_rate": 1.6049877917848948e-05, + "loss": 0.282, + "num_input_tokens_seen": 6685160, + "step": 24985 + }, + { + "epoch": 6.5504587155963305, + "grad_norm": 1.1820147037506104, + "learning_rate": 1.6039199639055755e-05, + "loss": 0.3244, + "num_input_tokens_seen": 6687512, + "step": 24990 + }, + { + "epoch": 6.551769331585845, + "grad_norm": 0.8672847747802734, + "learning_rate": 1.602852323574259e-05, + "loss": 0.3361, + "num_input_tokens_seen": 6689320, + "step": 24995 + }, + { + "epoch": 6.553079947575361, + "grad_norm": 0.8800460696220398, + "learning_rate": 1.6017848710144006e-05, + "loss": 0.1413, + "num_input_tokens_seen": 6690920, + "step": 25000 + }, + { + "epoch": 6.554390563564875, + "grad_norm": 1.3836723566055298, + "learning_rate": 1.600717606449417e-05, + "loss": 0.2531, + "num_input_tokens_seen": 6692088, + "step": 25005 + }, + { + "epoch": 6.555701179554391, + "grad_norm": 1.2903796434402466, + "learning_rate": 1.5996505301026848e-05, + "loss": 0.2635, + "num_input_tokens_seen": 6693288, + "step": 25010 + }, + { + "epoch": 6.557011795543906, + "grad_norm": 1.4058500528335571, + "learning_rate": 1.598583642197541e-05, + "loss": 0.2033, + "num_input_tokens_seen": 6694632, + "step": 25015 + }, + { + "epoch": 6.558322411533421, + "grad_norm": 1.1224068403244019, + "learning_rate": 1.5975169429572828e-05, + "loss": 0.0767, + "num_input_tokens_seen": 6695576, + "step": 25020 + }, + { + "epoch": 6.559633027522936, + "grad_norm": 2.795649290084839, + "learning_rate": 1.5964504326051693e-05, + "loss": 0.2238, + "num_input_tokens_seen": 6696536, + "step": 25025 + }, + { + "epoch": 6.56094364351245, + "grad_norm": 2.874486207962036, + "learning_rate": 1.595384111364419e-05, + "loss": 0.1756, + "num_input_tokens_seen": 6697960, + "step": 25030 + }, + { + "epoch": 6.562254259501966, + "grad_norm": 1.7269364595413208, + "learning_rate": 1.594317979458212e-05, + "loss": 0.2766, + "num_input_tokens_seen": 6699560, + "step": 25035 + }, + { + "epoch": 6.5635648754914815, + "grad_norm": 1.6774133443832397, + "learning_rate": 1.593252037109686e-05, + "loss": 0.1605, + "num_input_tokens_seen": 6700632, + "step": 25040 + }, + { + "epoch": 6.564875491480996, + "grad_norm": 2.1014010906219482, + "learning_rate": 1.5921862845419417e-05, + "loss": 0.259, + "num_input_tokens_seen": 6702312, + "step": 25045 + }, + { + "epoch": 6.566186107470511, + "grad_norm": 1.5886547565460205, + "learning_rate": 1.5911207219780398e-05, + "loss": 0.2792, + "num_input_tokens_seen": 6703816, + "step": 25050 + }, + { + "epoch": 6.567496723460026, + "grad_norm": 1.4014064073562622, + "learning_rate": 1.5900553496410004e-05, + "loss": 0.193, + "num_input_tokens_seen": 6705256, + "step": 25055 + }, + { + "epoch": 6.568807339449541, + "grad_norm": 2.4788522720336914, + "learning_rate": 1.5889901677538037e-05, + "loss": 0.257, + "num_input_tokens_seen": 6706408, + "step": 25060 + }, + { + "epoch": 6.5701179554390565, + "grad_norm": 2.300168514251709, + "learning_rate": 1.5879251765393914e-05, + "loss": 0.293, + "num_input_tokens_seen": 6707448, + "step": 25065 + }, + { + "epoch": 6.571428571428571, + "grad_norm": 1.2175915241241455, + "learning_rate": 1.5868603762206637e-05, + "loss": 0.2617, + "num_input_tokens_seen": 6708648, + "step": 25070 + }, + { + "epoch": 6.572739187418087, + "grad_norm": 1.4089831113815308, + "learning_rate": 1.5857957670204816e-05, + "loss": 0.2393, + "num_input_tokens_seen": 6709928, + "step": 25075 + }, + { + "epoch": 6.574049803407601, + "grad_norm": 2.0657432079315186, + "learning_rate": 1.5847313491616654e-05, + "loss": 0.2442, + "num_input_tokens_seen": 6711816, + "step": 25080 + }, + { + "epoch": 6.575360419397117, + "grad_norm": 1.2722417116165161, + "learning_rate": 1.5836671228669962e-05, + "loss": 0.2804, + "num_input_tokens_seen": 6713128, + "step": 25085 + }, + { + "epoch": 6.576671035386632, + "grad_norm": 2.697157621383667, + "learning_rate": 1.582603088359216e-05, + "loss": 0.1523, + "num_input_tokens_seen": 6714776, + "step": 25090 + }, + { + "epoch": 6.577981651376147, + "grad_norm": 3.338775873184204, + "learning_rate": 1.581539245861023e-05, + "loss": 0.2949, + "num_input_tokens_seen": 6715672, + "step": 25095 + }, + { + "epoch": 6.579292267365662, + "grad_norm": 0.971733808517456, + "learning_rate": 1.5804755955950785e-05, + "loss": 0.236, + "num_input_tokens_seen": 6717016, + "step": 25100 + }, + { + "epoch": 6.580602883355177, + "grad_norm": 1.850319266319275, + "learning_rate": 1.579412137784004e-05, + "loss": 0.1588, + "num_input_tokens_seen": 6718104, + "step": 25105 + }, + { + "epoch": 6.581913499344692, + "grad_norm": 1.2220367193222046, + "learning_rate": 1.578348872650378e-05, + "loss": 0.1718, + "num_input_tokens_seen": 6719064, + "step": 25110 + }, + { + "epoch": 6.5832241153342075, + "grad_norm": 3.782280683517456, + "learning_rate": 1.5772858004167407e-05, + "loss": 0.3696, + "num_input_tokens_seen": 6720376, + "step": 25115 + }, + { + "epoch": 6.584534731323722, + "grad_norm": 1.4720332622528076, + "learning_rate": 1.5762229213055914e-05, + "loss": 0.2121, + "num_input_tokens_seen": 6721784, + "step": 25120 + }, + { + "epoch": 6.585845347313237, + "grad_norm": 1.2865045070648193, + "learning_rate": 1.5751602355393878e-05, + "loss": 0.2051, + "num_input_tokens_seen": 6723192, + "step": 25125 + }, + { + "epoch": 6.587155963302752, + "grad_norm": 3.0794429779052734, + "learning_rate": 1.5740977433405496e-05, + "loss": 0.1762, + "num_input_tokens_seen": 6724392, + "step": 25130 + }, + { + "epoch": 6.588466579292267, + "grad_norm": 1.6811219453811646, + "learning_rate": 1.5730354449314534e-05, + "loss": 0.2599, + "num_input_tokens_seen": 6725672, + "step": 25135 + }, + { + "epoch": 6.5897771952817825, + "grad_norm": 1.3485167026519775, + "learning_rate": 1.5719733405344373e-05, + "loss": 0.2129, + "num_input_tokens_seen": 6727288, + "step": 25140 + }, + { + "epoch": 6.591087811271297, + "grad_norm": 2.6439456939697266, + "learning_rate": 1.570911430371797e-05, + "loss": 0.3603, + "num_input_tokens_seen": 6728744, + "step": 25145 + }, + { + "epoch": 6.592398427260813, + "grad_norm": 1.5438711643218994, + "learning_rate": 1.5698497146657897e-05, + "loss": 0.3268, + "num_input_tokens_seen": 6730040, + "step": 25150 + }, + { + "epoch": 6.593709043250327, + "grad_norm": 1.1431704759597778, + "learning_rate": 1.5687881936386294e-05, + "loss": 0.3123, + "num_input_tokens_seen": 6731592, + "step": 25155 + }, + { + "epoch": 6.595019659239843, + "grad_norm": 1.727890968322754, + "learning_rate": 1.5677268675124902e-05, + "loss": 0.3073, + "num_input_tokens_seen": 6733064, + "step": 25160 + }, + { + "epoch": 6.5963302752293576, + "grad_norm": 1.9698318243026733, + "learning_rate": 1.5666657365095083e-05, + "loss": 0.1348, + "num_input_tokens_seen": 6734216, + "step": 25165 + }, + { + "epoch": 6.597640891218873, + "grad_norm": 1.8727293014526367, + "learning_rate": 1.565604800851774e-05, + "loss": 0.1406, + "num_input_tokens_seen": 6735224, + "step": 25170 + }, + { + "epoch": 6.598951507208388, + "grad_norm": 1.6401684284210205, + "learning_rate": 1.5645440607613413e-05, + "loss": 0.2198, + "num_input_tokens_seen": 6736728, + "step": 25175 + }, + { + "epoch": 6.600262123197903, + "grad_norm": 0.8183385729789734, + "learning_rate": 1.56348351646022e-05, + "loss": 0.2471, + "num_input_tokens_seen": 6737720, + "step": 25180 + }, + { + "epoch": 6.601572739187418, + "grad_norm": 2.2312352657318115, + "learning_rate": 1.5624231681703812e-05, + "loss": 0.2715, + "num_input_tokens_seen": 6739016, + "step": 25185 + }, + { + "epoch": 6.6028833551769335, + "grad_norm": 1.1644858121871948, + "learning_rate": 1.5613630161137523e-05, + "loss": 0.1881, + "num_input_tokens_seen": 6740328, + "step": 25190 + }, + { + "epoch": 6.604193971166448, + "grad_norm": 2.3056912422180176, + "learning_rate": 1.5603030605122226e-05, + "loss": 0.1773, + "num_input_tokens_seen": 6741384, + "step": 25195 + }, + { + "epoch": 6.605504587155964, + "grad_norm": 1.964885950088501, + "learning_rate": 1.5592433015876383e-05, + "loss": 0.3004, + "num_input_tokens_seen": 6742744, + "step": 25200 + }, + { + "epoch": 6.606815203145478, + "grad_norm": 2.1727428436279297, + "learning_rate": 1.558183739561806e-05, + "loss": 0.4181, + "num_input_tokens_seen": 6746088, + "step": 25205 + }, + { + "epoch": 6.608125819134994, + "grad_norm": 1.992500901222229, + "learning_rate": 1.55712437465649e-05, + "loss": 0.1986, + "num_input_tokens_seen": 6747784, + "step": 25210 + }, + { + "epoch": 6.6094364351245085, + "grad_norm": 2.276388168334961, + "learning_rate": 1.5560652070934117e-05, + "loss": 0.4114, + "num_input_tokens_seen": 6748920, + "step": 25215 + }, + { + "epoch": 6.610747051114023, + "grad_norm": 2.2602717876434326, + "learning_rate": 1.5550062370942556e-05, + "loss": 0.2538, + "num_input_tokens_seen": 6750584, + "step": 25220 + }, + { + "epoch": 6.612057667103539, + "grad_norm": 0.9430841207504272, + "learning_rate": 1.553947464880662e-05, + "loss": 0.2177, + "num_input_tokens_seen": 6752184, + "step": 25225 + }, + { + "epoch": 6.613368283093053, + "grad_norm": 2.3264336585998535, + "learning_rate": 1.5528888906742285e-05, + "loss": 0.4218, + "num_input_tokens_seen": 6753480, + "step": 25230 + }, + { + "epoch": 6.614678899082569, + "grad_norm": 1.5094292163848877, + "learning_rate": 1.5518305146965133e-05, + "loss": 0.2771, + "num_input_tokens_seen": 6755032, + "step": 25235 + }, + { + "epoch": 6.6159895150720835, + "grad_norm": 1.7461384534835815, + "learning_rate": 1.5507723371690336e-05, + "loss": 0.15, + "num_input_tokens_seen": 6756680, + "step": 25240 + }, + { + "epoch": 6.617300131061599, + "grad_norm": 2.8417112827301025, + "learning_rate": 1.5497143583132638e-05, + "loss": 0.2448, + "num_input_tokens_seen": 6757896, + "step": 25245 + }, + { + "epoch": 6.618610747051114, + "grad_norm": 1.1694029569625854, + "learning_rate": 1.5486565783506367e-05, + "loss": 0.319, + "num_input_tokens_seen": 6759800, + "step": 25250 + }, + { + "epoch": 6.619921363040629, + "grad_norm": 1.7512816190719604, + "learning_rate": 1.547598997502544e-05, + "loss": 0.4445, + "num_input_tokens_seen": 6761096, + "step": 25255 + }, + { + "epoch": 6.621231979030144, + "grad_norm": 0.8481571674346924, + "learning_rate": 1.546541615990335e-05, + "loss": 0.2197, + "num_input_tokens_seen": 6762328, + "step": 25260 + }, + { + "epoch": 6.6225425950196595, + "grad_norm": 2.8028011322021484, + "learning_rate": 1.545484434035319e-05, + "loss": 0.2157, + "num_input_tokens_seen": 6763496, + "step": 25265 + }, + { + "epoch": 6.623853211009174, + "grad_norm": 1.9932329654693604, + "learning_rate": 1.5444274518587617e-05, + "loss": 0.3056, + "num_input_tokens_seen": 6764904, + "step": 25270 + }, + { + "epoch": 6.62516382699869, + "grad_norm": 1.6641793251037598, + "learning_rate": 1.543370669681886e-05, + "loss": 0.1993, + "num_input_tokens_seen": 6766168, + "step": 25275 + }, + { + "epoch": 6.626474442988204, + "grad_norm": 2.0316174030303955, + "learning_rate": 1.5423140877258784e-05, + "loss": 0.2854, + "num_input_tokens_seen": 6767272, + "step": 25280 + }, + { + "epoch": 6.62778505897772, + "grad_norm": 1.6674920320510864, + "learning_rate": 1.5412577062118763e-05, + "loss": 0.3645, + "num_input_tokens_seen": 6768456, + "step": 25285 + }, + { + "epoch": 6.6290956749672345, + "grad_norm": 1.3388257026672363, + "learning_rate": 1.5402015253609796e-05, + "loss": 0.238, + "num_input_tokens_seen": 6769800, + "step": 25290 + }, + { + "epoch": 6.63040629095675, + "grad_norm": 2.4748826026916504, + "learning_rate": 1.539145545394246e-05, + "loss": 0.1837, + "num_input_tokens_seen": 6771112, + "step": 25295 + }, + { + "epoch": 6.631716906946265, + "grad_norm": 2.204383373260498, + "learning_rate": 1.538089766532689e-05, + "loss": 0.307, + "num_input_tokens_seen": 6772840, + "step": 25300 + }, + { + "epoch": 6.63302752293578, + "grad_norm": 1.8661317825317383, + "learning_rate": 1.537034188997282e-05, + "loss": 0.1975, + "num_input_tokens_seen": 6774008, + "step": 25305 + }, + { + "epoch": 6.634338138925295, + "grad_norm": 2.27575945854187, + "learning_rate": 1.5359788130089558e-05, + "loss": 0.3276, + "num_input_tokens_seen": 6775336, + "step": 25310 + }, + { + "epoch": 6.6356487549148095, + "grad_norm": 1.0249584913253784, + "learning_rate": 1.5349236387885976e-05, + "loss": 0.2482, + "num_input_tokens_seen": 6776872, + "step": 25315 + }, + { + "epoch": 6.636959370904325, + "grad_norm": 1.3294677734375, + "learning_rate": 1.5338686665570546e-05, + "loss": 0.2453, + "num_input_tokens_seen": 6778040, + "step": 25320 + }, + { + "epoch": 6.63826998689384, + "grad_norm": 1.2166531085968018, + "learning_rate": 1.5328138965351307e-05, + "loss": 0.2636, + "num_input_tokens_seen": 6779864, + "step": 25325 + }, + { + "epoch": 6.639580602883355, + "grad_norm": 1.564656138420105, + "learning_rate": 1.5317593289435855e-05, + "loss": 0.2961, + "num_input_tokens_seen": 6781800, + "step": 25330 + }, + { + "epoch": 6.64089121887287, + "grad_norm": 1.706754207611084, + "learning_rate": 1.530704964003141e-05, + "loss": 0.3089, + "num_input_tokens_seen": 6783208, + "step": 25335 + }, + { + "epoch": 6.6422018348623855, + "grad_norm": 1.4308342933654785, + "learning_rate": 1.529650801934473e-05, + "loss": 0.2658, + "num_input_tokens_seen": 6784664, + "step": 25340 + }, + { + "epoch": 6.6435124508519, + "grad_norm": 1.5188485383987427, + "learning_rate": 1.5285968429582153e-05, + "loss": 0.2061, + "num_input_tokens_seen": 6786056, + "step": 25345 + }, + { + "epoch": 6.644823066841416, + "grad_norm": 1.4032869338989258, + "learning_rate": 1.52754308729496e-05, + "loss": 0.3251, + "num_input_tokens_seen": 6787656, + "step": 25350 + }, + { + "epoch": 6.64613368283093, + "grad_norm": 3.127634048461914, + "learning_rate": 1.5264895351652555e-05, + "loss": 0.3597, + "num_input_tokens_seen": 6788728, + "step": 25355 + }, + { + "epoch": 6.647444298820446, + "grad_norm": 1.2709481716156006, + "learning_rate": 1.52543618678961e-05, + "loss": 0.151, + "num_input_tokens_seen": 6789832, + "step": 25360 + }, + { + "epoch": 6.6487549148099605, + "grad_norm": 1.2666828632354736, + "learning_rate": 1.5243830423884867e-05, + "loss": 0.1814, + "num_input_tokens_seen": 6791368, + "step": 25365 + }, + { + "epoch": 6.650065530799476, + "grad_norm": 3.605217218399048, + "learning_rate": 1.5233301021823065e-05, + "loss": 0.3249, + "num_input_tokens_seen": 6792648, + "step": 25370 + }, + { + "epoch": 6.651376146788991, + "grad_norm": 2.8100719451904297, + "learning_rate": 1.5222773663914486e-05, + "loss": 0.1692, + "num_input_tokens_seen": 6794120, + "step": 25375 + }, + { + "epoch": 6.652686762778506, + "grad_norm": 1.331608533859253, + "learning_rate": 1.5212248352362482e-05, + "loss": 0.3532, + "num_input_tokens_seen": 6795240, + "step": 25380 + }, + { + "epoch": 6.653997378768021, + "grad_norm": 1.5856058597564697, + "learning_rate": 1.5201725089369991e-05, + "loss": 0.1986, + "num_input_tokens_seen": 6796632, + "step": 25385 + }, + { + "epoch": 6.655307994757536, + "grad_norm": 0.98968106508255, + "learning_rate": 1.5191203877139498e-05, + "loss": 0.1829, + "num_input_tokens_seen": 6798120, + "step": 25390 + }, + { + "epoch": 6.656618610747051, + "grad_norm": 0.9531396627426147, + "learning_rate": 1.5180684717873089e-05, + "loss": 0.3465, + "num_input_tokens_seen": 6799784, + "step": 25395 + }, + { + "epoch": 6.657929226736567, + "grad_norm": 2.4640185832977295, + "learning_rate": 1.5170167613772412e-05, + "loss": 0.2431, + "num_input_tokens_seen": 6800952, + "step": 25400 + }, + { + "epoch": 6.659239842726081, + "grad_norm": 2.8397979736328125, + "learning_rate": 1.5159652567038663e-05, + "loss": 0.3419, + "num_input_tokens_seen": 6802504, + "step": 25405 + }, + { + "epoch": 6.660550458715596, + "grad_norm": 1.4146803617477417, + "learning_rate": 1.5149139579872629e-05, + "loss": 0.312, + "num_input_tokens_seen": 6803880, + "step": 25410 + }, + { + "epoch": 6.6618610747051115, + "grad_norm": 1.338234782218933, + "learning_rate": 1.5138628654474656e-05, + "loss": 0.2962, + "num_input_tokens_seen": 6805176, + "step": 25415 + }, + { + "epoch": 6.663171690694626, + "grad_norm": 2.252063274383545, + "learning_rate": 1.5128119793044677e-05, + "loss": 0.2088, + "num_input_tokens_seen": 6806456, + "step": 25420 + }, + { + "epoch": 6.664482306684142, + "grad_norm": 3.4062581062316895, + "learning_rate": 1.511761299778216e-05, + "loss": 0.2529, + "num_input_tokens_seen": 6807592, + "step": 25425 + }, + { + "epoch": 6.665792922673656, + "grad_norm": 1.7704654932022095, + "learning_rate": 1.5107108270886166e-05, + "loss": 0.3242, + "num_input_tokens_seen": 6809448, + "step": 25430 + }, + { + "epoch": 6.667103538663172, + "grad_norm": 1.443550944328308, + "learning_rate": 1.509660561455532e-05, + "loss": 0.2938, + "num_input_tokens_seen": 6810520, + "step": 25435 + }, + { + "epoch": 6.6684141546526865, + "grad_norm": 2.525733232498169, + "learning_rate": 1.5086105030987807e-05, + "loss": 0.2913, + "num_input_tokens_seen": 6811656, + "step": 25440 + }, + { + "epoch": 6.669724770642202, + "grad_norm": 1.4297454357147217, + "learning_rate": 1.5075606522381375e-05, + "loss": 0.1949, + "num_input_tokens_seen": 6812600, + "step": 25445 + }, + { + "epoch": 6.671035386631717, + "grad_norm": 0.6458961367607117, + "learning_rate": 1.5065110090933354e-05, + "loss": 0.3124, + "num_input_tokens_seen": 6813912, + "step": 25450 + }, + { + "epoch": 6.672346002621232, + "grad_norm": 0.9936894774436951, + "learning_rate": 1.5054615738840627e-05, + "loss": 0.2779, + "num_input_tokens_seen": 6814904, + "step": 25455 + }, + { + "epoch": 6.673656618610747, + "grad_norm": 1.4605096578598022, + "learning_rate": 1.5044123468299648e-05, + "loss": 0.3012, + "num_input_tokens_seen": 6815992, + "step": 25460 + }, + { + "epoch": 6.674967234600262, + "grad_norm": 1.244712233543396, + "learning_rate": 1.5033633281506416e-05, + "loss": 0.3217, + "num_input_tokens_seen": 6817640, + "step": 25465 + }, + { + "epoch": 6.676277850589777, + "grad_norm": 2.4847683906555176, + "learning_rate": 1.5023145180656517e-05, + "loss": 0.3114, + "num_input_tokens_seen": 6819288, + "step": 25470 + }, + { + "epoch": 6.677588466579293, + "grad_norm": 2.1611454486846924, + "learning_rate": 1.5012659167945097e-05, + "loss": 0.245, + "num_input_tokens_seen": 6820440, + "step": 25475 + }, + { + "epoch": 6.678899082568807, + "grad_norm": 0.961096465587616, + "learning_rate": 1.5002175245566857e-05, + "loss": 0.1898, + "num_input_tokens_seen": 6822104, + "step": 25480 + }, + { + "epoch": 6.680209698558322, + "grad_norm": 1.976837396621704, + "learning_rate": 1.4991693415716062e-05, + "loss": 0.368, + "num_input_tokens_seen": 6823208, + "step": 25485 + }, + { + "epoch": 6.6815203145478375, + "grad_norm": 1.7837021350860596, + "learning_rate": 1.498121368058654e-05, + "loss": 0.2721, + "num_input_tokens_seen": 6824872, + "step": 25490 + }, + { + "epoch": 6.682830930537353, + "grad_norm": 1.855972409248352, + "learning_rate": 1.4970736042371685e-05, + "loss": 0.222, + "num_input_tokens_seen": 6826040, + "step": 25495 + }, + { + "epoch": 6.684141546526868, + "grad_norm": 2.407021999359131, + "learning_rate": 1.4960260503264445e-05, + "loss": 0.2761, + "num_input_tokens_seen": 6827400, + "step": 25500 + }, + { + "epoch": 6.685452162516382, + "grad_norm": 1.3238797187805176, + "learning_rate": 1.494978706545733e-05, + "loss": 0.1884, + "num_input_tokens_seen": 6828760, + "step": 25505 + }, + { + "epoch": 6.686762778505898, + "grad_norm": 1.738951325416565, + "learning_rate": 1.4939315731142423e-05, + "loss": 0.1932, + "num_input_tokens_seen": 6830296, + "step": 25510 + }, + { + "epoch": 6.6880733944954125, + "grad_norm": 1.4242318868637085, + "learning_rate": 1.4928846502511353e-05, + "loss": 0.3092, + "num_input_tokens_seen": 6832152, + "step": 25515 + }, + { + "epoch": 6.689384010484928, + "grad_norm": 1.748123288154602, + "learning_rate": 1.4918379381755303e-05, + "loss": 0.2365, + "num_input_tokens_seen": 6833288, + "step": 25520 + }, + { + "epoch": 6.690694626474443, + "grad_norm": 0.7206429839134216, + "learning_rate": 1.4907914371065032e-05, + "loss": 0.11, + "num_input_tokens_seen": 6834792, + "step": 25525 + }, + { + "epoch": 6.692005242463958, + "grad_norm": 2.8795828819274902, + "learning_rate": 1.4897451472630844e-05, + "loss": 0.3807, + "num_input_tokens_seen": 6836360, + "step": 25530 + }, + { + "epoch": 6.693315858453473, + "grad_norm": 0.6078527569770813, + "learning_rate": 1.4886990688642605e-05, + "loss": 0.1987, + "num_input_tokens_seen": 6837448, + "step": 25535 + }, + { + "epoch": 6.694626474442988, + "grad_norm": 1.8639463186264038, + "learning_rate": 1.487653202128975e-05, + "loss": 0.3264, + "num_input_tokens_seen": 6838632, + "step": 25540 + }, + { + "epoch": 6.695937090432503, + "grad_norm": 4.406266212463379, + "learning_rate": 1.4866075472761243e-05, + "loss": 0.2492, + "num_input_tokens_seen": 6839720, + "step": 25545 + }, + { + "epoch": 6.697247706422019, + "grad_norm": 2.091315746307373, + "learning_rate": 1.485562104524563e-05, + "loss": 0.2556, + "num_input_tokens_seen": 6840856, + "step": 25550 + }, + { + "epoch": 6.698558322411533, + "grad_norm": 1.3506547212600708, + "learning_rate": 1.4845168740931004e-05, + "loss": 0.2303, + "num_input_tokens_seen": 6842088, + "step": 25555 + }, + { + "epoch": 6.699868938401049, + "grad_norm": 0.9270949959754944, + "learning_rate": 1.4834718562005007e-05, + "loss": 0.155, + "num_input_tokens_seen": 6843528, + "step": 25560 + }, + { + "epoch": 6.7011795543905635, + "grad_norm": 2.441312789916992, + "learning_rate": 1.4824270510654859e-05, + "loss": 0.229, + "num_input_tokens_seen": 6844776, + "step": 25565 + }, + { + "epoch": 6.702490170380079, + "grad_norm": 0.6789252758026123, + "learning_rate": 1.4813824589067309e-05, + "loss": 0.1508, + "num_input_tokens_seen": 6846216, + "step": 25570 + }, + { + "epoch": 6.703800786369594, + "grad_norm": 5.260751247406006, + "learning_rate": 1.4803380799428679e-05, + "loss": 0.2156, + "num_input_tokens_seen": 6847224, + "step": 25575 + }, + { + "epoch": 6.705111402359108, + "grad_norm": 2.036958932876587, + "learning_rate": 1.4792939143924822e-05, + "loss": 0.2884, + "num_input_tokens_seen": 6848536, + "step": 25580 + }, + { + "epoch": 6.706422018348624, + "grad_norm": 2.268080472946167, + "learning_rate": 1.478249962474117e-05, + "loss": 0.3374, + "num_input_tokens_seen": 6849672, + "step": 25585 + }, + { + "epoch": 6.707732634338139, + "grad_norm": 1.418733835220337, + "learning_rate": 1.4772062244062687e-05, + "loss": 0.2673, + "num_input_tokens_seen": 6850824, + "step": 25590 + }, + { + "epoch": 6.709043250327654, + "grad_norm": 2.417398452758789, + "learning_rate": 1.4761627004073905e-05, + "loss": 0.2804, + "num_input_tokens_seen": 6852072, + "step": 25595 + }, + { + "epoch": 6.710353866317169, + "grad_norm": 2.7518770694732666, + "learning_rate": 1.4751193906958907e-05, + "loss": 0.2672, + "num_input_tokens_seen": 6853240, + "step": 25600 + }, + { + "epoch": 6.711664482306684, + "grad_norm": 2.5876970291137695, + "learning_rate": 1.474076295490131e-05, + "loss": 0.2083, + "num_input_tokens_seen": 6854840, + "step": 25605 + }, + { + "epoch": 6.712975098296199, + "grad_norm": 2.794415235519409, + "learning_rate": 1.4730334150084302e-05, + "loss": 0.1712, + "num_input_tokens_seen": 6855848, + "step": 25610 + }, + { + "epoch": 6.714285714285714, + "grad_norm": 1.1335200071334839, + "learning_rate": 1.4719907494690604e-05, + "loss": 0.2375, + "num_input_tokens_seen": 6857320, + "step": 25615 + }, + { + "epoch": 6.715596330275229, + "grad_norm": 2.3045339584350586, + "learning_rate": 1.470948299090251e-05, + "loss": 0.1912, + "num_input_tokens_seen": 6858568, + "step": 25620 + }, + { + "epoch": 6.716906946264745, + "grad_norm": 1.794236421585083, + "learning_rate": 1.4699060640901845e-05, + "loss": 0.1149, + "num_input_tokens_seen": 6859912, + "step": 25625 + }, + { + "epoch": 6.718217562254259, + "grad_norm": 1.3602977991104126, + "learning_rate": 1.4688640446869988e-05, + "loss": 0.1465, + "num_input_tokens_seen": 6860760, + "step": 25630 + }, + { + "epoch": 6.719528178243775, + "grad_norm": 4.028629302978516, + "learning_rate": 1.467822241098788e-05, + "loss": 0.2182, + "num_input_tokens_seen": 6861928, + "step": 25635 + }, + { + "epoch": 6.7208387942332894, + "grad_norm": 2.235368251800537, + "learning_rate": 1.4667806535435974e-05, + "loss": 0.142, + "num_input_tokens_seen": 6864136, + "step": 25640 + }, + { + "epoch": 6.722149410222805, + "grad_norm": 1.2360671758651733, + "learning_rate": 1.4657392822394312e-05, + "loss": 0.1916, + "num_input_tokens_seen": 6865480, + "step": 25645 + }, + { + "epoch": 6.72346002621232, + "grad_norm": 1.5426806211471558, + "learning_rate": 1.464698127404246e-05, + "loss": 0.2111, + "num_input_tokens_seen": 6866888, + "step": 25650 + }, + { + "epoch": 6.724770642201835, + "grad_norm": 1.698303461074829, + "learning_rate": 1.4636571892559547e-05, + "loss": 0.3076, + "num_input_tokens_seen": 6868248, + "step": 25655 + }, + { + "epoch": 6.72608125819135, + "grad_norm": 0.8018031120300293, + "learning_rate": 1.4626164680124221e-05, + "loss": 0.2203, + "num_input_tokens_seen": 6869352, + "step": 25660 + }, + { + "epoch": 6.727391874180865, + "grad_norm": 2.6353518962860107, + "learning_rate": 1.4615759638914705e-05, + "loss": 0.2508, + "num_input_tokens_seen": 6870440, + "step": 25665 + }, + { + "epoch": 6.72870249017038, + "grad_norm": 2.555394411087036, + "learning_rate": 1.4605356771108753e-05, + "loss": 0.3034, + "num_input_tokens_seen": 6871704, + "step": 25670 + }, + { + "epoch": 6.730013106159895, + "grad_norm": 1.2971445322036743, + "learning_rate": 1.4594956078883665e-05, + "loss": 0.2152, + "num_input_tokens_seen": 6873208, + "step": 25675 + }, + { + "epoch": 6.73132372214941, + "grad_norm": 1.9627256393432617, + "learning_rate": 1.4584557564416296e-05, + "loss": 0.2613, + "num_input_tokens_seen": 6874616, + "step": 25680 + }, + { + "epoch": 6.732634338138926, + "grad_norm": 2.392857313156128, + "learning_rate": 1.457416122988303e-05, + "loss": 0.2078, + "num_input_tokens_seen": 6875672, + "step": 25685 + }, + { + "epoch": 6.73394495412844, + "grad_norm": 1.207236647605896, + "learning_rate": 1.4563767077459805e-05, + "loss": 0.2112, + "num_input_tokens_seen": 6877064, + "step": 25690 + }, + { + "epoch": 6.735255570117955, + "grad_norm": 3.47766375541687, + "learning_rate": 1.4553375109322098e-05, + "loss": 0.2737, + "num_input_tokens_seen": 6878328, + "step": 25695 + }, + { + "epoch": 6.736566186107471, + "grad_norm": 1.648040533065796, + "learning_rate": 1.4542985327644931e-05, + "loss": 0.2189, + "num_input_tokens_seen": 6879528, + "step": 25700 + }, + { + "epoch": 6.737876802096985, + "grad_norm": 1.4637372493743896, + "learning_rate": 1.4532597734602877e-05, + "loss": 0.303, + "num_input_tokens_seen": 6880920, + "step": 25705 + }, + { + "epoch": 6.739187418086501, + "grad_norm": 2.285416841506958, + "learning_rate": 1.4522212332370019e-05, + "loss": 0.2301, + "num_input_tokens_seen": 6882344, + "step": 25710 + }, + { + "epoch": 6.740498034076015, + "grad_norm": 2.549225330352783, + "learning_rate": 1.4511829123120018e-05, + "loss": 0.4123, + "num_input_tokens_seen": 6883624, + "step": 25715 + }, + { + "epoch": 6.741808650065531, + "grad_norm": 1.7366043329238892, + "learning_rate": 1.4501448109026055e-05, + "loss": 0.2803, + "num_input_tokens_seen": 6885432, + "step": 25720 + }, + { + "epoch": 6.743119266055046, + "grad_norm": 1.1062172651290894, + "learning_rate": 1.4491069292260868e-05, + "loss": 0.1951, + "num_input_tokens_seen": 6887480, + "step": 25725 + }, + { + "epoch": 6.744429882044561, + "grad_norm": 7.518061637878418, + "learning_rate": 1.448069267499671e-05, + "loss": 0.2426, + "num_input_tokens_seen": 6888728, + "step": 25730 + }, + { + "epoch": 6.745740498034076, + "grad_norm": 1.8967020511627197, + "learning_rate": 1.4470318259405407e-05, + "loss": 0.5907, + "num_input_tokens_seen": 6889848, + "step": 25735 + }, + { + "epoch": 6.747051114023591, + "grad_norm": 1.0018972158432007, + "learning_rate": 1.4459946047658306e-05, + "loss": 0.1577, + "num_input_tokens_seen": 6891016, + "step": 25740 + }, + { + "epoch": 6.748361730013106, + "grad_norm": 1.5092134475708008, + "learning_rate": 1.444957604192628e-05, + "loss": 0.127, + "num_input_tokens_seen": 6892472, + "step": 25745 + }, + { + "epoch": 6.749672346002622, + "grad_norm": 0.4584087133407593, + "learning_rate": 1.4439208244379757e-05, + "loss": 0.2941, + "num_input_tokens_seen": 6893912, + "step": 25750 + }, + { + "epoch": 6.750982961992136, + "grad_norm": 1.6184319257736206, + "learning_rate": 1.4428842657188701e-05, + "loss": 0.2574, + "num_input_tokens_seen": 6894920, + "step": 25755 + }, + { + "epoch": 6.752293577981652, + "grad_norm": 1.580262541770935, + "learning_rate": 1.441847928252261e-05, + "loss": 0.2566, + "num_input_tokens_seen": 6896344, + "step": 25760 + }, + { + "epoch": 6.753604193971166, + "grad_norm": 1.4148156642913818, + "learning_rate": 1.4408118122550524e-05, + "loss": 0.3483, + "num_input_tokens_seen": 6898072, + "step": 25765 + }, + { + "epoch": 6.754914809960681, + "grad_norm": 1.3736826181411743, + "learning_rate": 1.4397759179441012e-05, + "loss": 0.442, + "num_input_tokens_seen": 6899304, + "step": 25770 + }, + { + "epoch": 6.756225425950197, + "grad_norm": 1.2735506296157837, + "learning_rate": 1.4387402455362181e-05, + "loss": 0.165, + "num_input_tokens_seen": 6901000, + "step": 25775 + }, + { + "epoch": 6.757536041939712, + "grad_norm": 1.6240298748016357, + "learning_rate": 1.4377047952481685e-05, + "loss": 0.2755, + "num_input_tokens_seen": 6902184, + "step": 25780 + }, + { + "epoch": 6.758846657929227, + "grad_norm": 2.7344954013824463, + "learning_rate": 1.4366695672966705e-05, + "loss": 0.2029, + "num_input_tokens_seen": 6903480, + "step": 25785 + }, + { + "epoch": 6.760157273918741, + "grad_norm": 2.061455488204956, + "learning_rate": 1.4356345618983927e-05, + "loss": 0.239, + "num_input_tokens_seen": 6904536, + "step": 25790 + }, + { + "epoch": 6.761467889908257, + "grad_norm": 2.8507871627807617, + "learning_rate": 1.434599779269963e-05, + "loss": 0.1752, + "num_input_tokens_seen": 6905768, + "step": 25795 + }, + { + "epoch": 6.762778505897772, + "grad_norm": 0.8657524585723877, + "learning_rate": 1.4335652196279587e-05, + "loss": 0.15, + "num_input_tokens_seen": 6907096, + "step": 25800 + }, + { + "epoch": 6.764089121887287, + "grad_norm": 3.2279839515686035, + "learning_rate": 1.4325308831889115e-05, + "loss": 0.2548, + "num_input_tokens_seen": 6908344, + "step": 25805 + }, + { + "epoch": 6.765399737876802, + "grad_norm": 1.983083724975586, + "learning_rate": 1.4314967701693055e-05, + "loss": 0.224, + "num_input_tokens_seen": 6909592, + "step": 25810 + }, + { + "epoch": 6.766710353866317, + "grad_norm": 1.5988945960998535, + "learning_rate": 1.4304628807855797e-05, + "loss": 0.1939, + "num_input_tokens_seen": 6910552, + "step": 25815 + }, + { + "epoch": 6.768020969855832, + "grad_norm": 2.020080804824829, + "learning_rate": 1.4294292152541258e-05, + "loss": 0.255, + "num_input_tokens_seen": 6911880, + "step": 25820 + }, + { + "epoch": 6.7693315858453476, + "grad_norm": 1.3382413387298584, + "learning_rate": 1.4283957737912862e-05, + "loss": 0.2539, + "num_input_tokens_seen": 6913176, + "step": 25825 + }, + { + "epoch": 6.770642201834862, + "grad_norm": 2.0145883560180664, + "learning_rate": 1.4273625566133596e-05, + "loss": 0.3494, + "num_input_tokens_seen": 6914424, + "step": 25830 + }, + { + "epoch": 6.771952817824378, + "grad_norm": 2.008516311645508, + "learning_rate": 1.426329563936597e-05, + "loss": 0.2599, + "num_input_tokens_seen": 6915896, + "step": 25835 + }, + { + "epoch": 6.773263433813892, + "grad_norm": 1.0743905305862427, + "learning_rate": 1.4252967959772013e-05, + "loss": 0.2254, + "num_input_tokens_seen": 6917672, + "step": 25840 + }, + { + "epoch": 6.774574049803408, + "grad_norm": 0.9012834429740906, + "learning_rate": 1.4242642529513278e-05, + "loss": 0.1519, + "num_input_tokens_seen": 6918968, + "step": 25845 + }, + { + "epoch": 6.775884665792923, + "grad_norm": 1.676035761833191, + "learning_rate": 1.4232319350750888e-05, + "loss": 0.2545, + "num_input_tokens_seen": 6920536, + "step": 25850 + }, + { + "epoch": 6.777195281782438, + "grad_norm": 1.4625710248947144, + "learning_rate": 1.422199842564545e-05, + "loss": 0.2209, + "num_input_tokens_seen": 6922056, + "step": 25855 + }, + { + "epoch": 6.778505897771953, + "grad_norm": 1.440205693244934, + "learning_rate": 1.4211679756357132e-05, + "loss": 0.2738, + "num_input_tokens_seen": 6923080, + "step": 25860 + }, + { + "epoch": 6.779816513761467, + "grad_norm": 1.2217859029769897, + "learning_rate": 1.4201363345045588e-05, + "loss": 0.2314, + "num_input_tokens_seen": 6924776, + "step": 25865 + }, + { + "epoch": 6.781127129750983, + "grad_norm": 2.1064724922180176, + "learning_rate": 1.4191049193870037e-05, + "loss": 0.2396, + "num_input_tokens_seen": 6926216, + "step": 25870 + }, + { + "epoch": 6.7824377457404985, + "grad_norm": 1.5648903846740723, + "learning_rate": 1.4180737304989216e-05, + "loss": 0.2452, + "num_input_tokens_seen": 6927432, + "step": 25875 + }, + { + "epoch": 6.783748361730013, + "grad_norm": 1.880053162574768, + "learning_rate": 1.4170427680561382e-05, + "loss": 0.2215, + "num_input_tokens_seen": 6929064, + "step": 25880 + }, + { + "epoch": 6.785058977719528, + "grad_norm": 1.1547431945800781, + "learning_rate": 1.4160120322744322e-05, + "loss": 0.2344, + "num_input_tokens_seen": 6930584, + "step": 25885 + }, + { + "epoch": 6.786369593709043, + "grad_norm": 1.767004132270813, + "learning_rate": 1.4149815233695346e-05, + "loss": 0.3393, + "num_input_tokens_seen": 6932312, + "step": 25890 + }, + { + "epoch": 6.787680209698558, + "grad_norm": 1.4603053331375122, + "learning_rate": 1.4139512415571298e-05, + "loss": 0.3432, + "num_input_tokens_seen": 6934152, + "step": 25895 + }, + { + "epoch": 6.7889908256880735, + "grad_norm": 1.833279013633728, + "learning_rate": 1.4129211870528534e-05, + "loss": 0.207, + "num_input_tokens_seen": 6935400, + "step": 25900 + }, + { + "epoch": 6.790301441677588, + "grad_norm": 1.2790592908859253, + "learning_rate": 1.411891360072294e-05, + "loss": 0.1564, + "num_input_tokens_seen": 6936424, + "step": 25905 + }, + { + "epoch": 6.791612057667104, + "grad_norm": 3.8724899291992188, + "learning_rate": 1.4108617608309932e-05, + "loss": 0.1957, + "num_input_tokens_seen": 6937624, + "step": 25910 + }, + { + "epoch": 6.792922673656618, + "grad_norm": 1.2409392595291138, + "learning_rate": 1.4098323895444437e-05, + "loss": 0.2361, + "num_input_tokens_seen": 6939032, + "step": 25915 + }, + { + "epoch": 6.794233289646134, + "grad_norm": 1.4805785417556763, + "learning_rate": 1.4088032464280914e-05, + "loss": 0.352, + "num_input_tokens_seen": 6940296, + "step": 25920 + }, + { + "epoch": 6.795543905635649, + "grad_norm": 1.9368220567703247, + "learning_rate": 1.4077743316973346e-05, + "loss": 0.2502, + "num_input_tokens_seen": 6942200, + "step": 25925 + }, + { + "epoch": 6.796854521625164, + "grad_norm": 1.993643045425415, + "learning_rate": 1.4067456455675226e-05, + "loss": 0.2461, + "num_input_tokens_seen": 6944472, + "step": 25930 + }, + { + "epoch": 6.798165137614679, + "grad_norm": 1.1009455919265747, + "learning_rate": 1.4057171882539583e-05, + "loss": 0.2642, + "num_input_tokens_seen": 6945944, + "step": 25935 + }, + { + "epoch": 6.799475753604194, + "grad_norm": 1.8125256299972534, + "learning_rate": 1.4046889599718962e-05, + "loss": 0.1961, + "num_input_tokens_seen": 6947096, + "step": 25940 + }, + { + "epoch": 6.800786369593709, + "grad_norm": 1.8061448335647583, + "learning_rate": 1.4036609609365414e-05, + "loss": 0.2583, + "num_input_tokens_seen": 6948360, + "step": 25945 + }, + { + "epoch": 6.8020969855832245, + "grad_norm": 2.179001808166504, + "learning_rate": 1.4026331913630536e-05, + "loss": 0.2488, + "num_input_tokens_seen": 6949576, + "step": 25950 + }, + { + "epoch": 6.803407601572739, + "grad_norm": 2.5903406143188477, + "learning_rate": 1.401605651466542e-05, + "loss": 0.1504, + "num_input_tokens_seen": 6950840, + "step": 25955 + }, + { + "epoch": 6.804718217562254, + "grad_norm": 2.219332218170166, + "learning_rate": 1.4005783414620694e-05, + "loss": 0.1805, + "num_input_tokens_seen": 6952296, + "step": 25960 + }, + { + "epoch": 6.806028833551769, + "grad_norm": 1.7607508897781372, + "learning_rate": 1.3995512615646509e-05, + "loss": 0.3162, + "num_input_tokens_seen": 6953512, + "step": 25965 + }, + { + "epoch": 6.807339449541285, + "grad_norm": 2.1444172859191895, + "learning_rate": 1.3985244119892518e-05, + "loss": 0.247, + "num_input_tokens_seen": 6954744, + "step": 25970 + }, + { + "epoch": 6.8086500655307995, + "grad_norm": 0.8672334551811218, + "learning_rate": 1.3974977929507913e-05, + "loss": 0.1788, + "num_input_tokens_seen": 6956280, + "step": 25975 + }, + { + "epoch": 6.809960681520314, + "grad_norm": 3.5141706466674805, + "learning_rate": 1.3964714046641366e-05, + "loss": 0.1871, + "num_input_tokens_seen": 6957304, + "step": 25980 + }, + { + "epoch": 6.81127129750983, + "grad_norm": 1.4606006145477295, + "learning_rate": 1.3954452473441104e-05, + "loss": 0.2493, + "num_input_tokens_seen": 6958296, + "step": 25985 + }, + { + "epoch": 6.812581913499344, + "grad_norm": 1.4185938835144043, + "learning_rate": 1.3944193212054851e-05, + "loss": 0.2981, + "num_input_tokens_seen": 6959400, + "step": 25990 + }, + { + "epoch": 6.81389252948886, + "grad_norm": 2.082223415374756, + "learning_rate": 1.393393626462986e-05, + "loss": 0.211, + "num_input_tokens_seen": 6960904, + "step": 25995 + }, + { + "epoch": 6.815203145478375, + "grad_norm": 1.855131983757019, + "learning_rate": 1.3923681633312885e-05, + "loss": 0.1558, + "num_input_tokens_seen": 6962120, + "step": 26000 + }, + { + "epoch": 6.81651376146789, + "grad_norm": 1.6895384788513184, + "learning_rate": 1.3913429320250203e-05, + "loss": 0.1931, + "num_input_tokens_seen": 6963160, + "step": 26005 + }, + { + "epoch": 6.817824377457405, + "grad_norm": 0.4953508675098419, + "learning_rate": 1.3903179327587612e-05, + "loss": 0.2885, + "num_input_tokens_seen": 6964696, + "step": 26010 + }, + { + "epoch": 6.81913499344692, + "grad_norm": 1.6412330865859985, + "learning_rate": 1.3892931657470415e-05, + "loss": 0.1899, + "num_input_tokens_seen": 6966152, + "step": 26015 + }, + { + "epoch": 6.820445609436435, + "grad_norm": 4.600361347198486, + "learning_rate": 1.3882686312043428e-05, + "loss": 0.317, + "num_input_tokens_seen": 6967320, + "step": 26020 + }, + { + "epoch": 6.8217562254259505, + "grad_norm": 3.3635427951812744, + "learning_rate": 1.3872443293450987e-05, + "loss": 0.261, + "num_input_tokens_seen": 6968440, + "step": 26025 + }, + { + "epoch": 6.823066841415465, + "grad_norm": 1.9717177152633667, + "learning_rate": 1.3862202603836938e-05, + "loss": 0.2756, + "num_input_tokens_seen": 6969800, + "step": 26030 + }, + { + "epoch": 6.824377457404981, + "grad_norm": 1.3314868211746216, + "learning_rate": 1.385196424534464e-05, + "loss": 0.2207, + "num_input_tokens_seen": 6971752, + "step": 26035 + }, + { + "epoch": 6.825688073394495, + "grad_norm": 4.088747024536133, + "learning_rate": 1.3841728220116961e-05, + "loss": 0.159, + "num_input_tokens_seen": 6972968, + "step": 26040 + }, + { + "epoch": 6.826998689384011, + "grad_norm": 1.4322189092636108, + "learning_rate": 1.3831494530296291e-05, + "loss": 0.1474, + "num_input_tokens_seen": 6974664, + "step": 26045 + }, + { + "epoch": 6.8283093053735255, + "grad_norm": 1.4270190000534058, + "learning_rate": 1.3821263178024512e-05, + "loss": 0.2071, + "num_input_tokens_seen": 6976152, + "step": 26050 + }, + { + "epoch": 6.82961992136304, + "grad_norm": 2.029226303100586, + "learning_rate": 1.3811034165443038e-05, + "loss": 0.2499, + "num_input_tokens_seen": 6977400, + "step": 26055 + }, + { + "epoch": 6.830930537352556, + "grad_norm": 2.3830690383911133, + "learning_rate": 1.3800807494692785e-05, + "loss": 0.3321, + "num_input_tokens_seen": 6978808, + "step": 26060 + }, + { + "epoch": 6.832241153342071, + "grad_norm": 1.9512676000595093, + "learning_rate": 1.3790583167914168e-05, + "loss": 0.2299, + "num_input_tokens_seen": 6980424, + "step": 26065 + }, + { + "epoch": 6.833551769331586, + "grad_norm": 3.2721340656280518, + "learning_rate": 1.3780361187247125e-05, + "loss": 0.1886, + "num_input_tokens_seen": 6981736, + "step": 26070 + }, + { + "epoch": 6.834862385321101, + "grad_norm": 1.3400087356567383, + "learning_rate": 1.377014155483109e-05, + "loss": 0.37, + "num_input_tokens_seen": 6983112, + "step": 26075 + }, + { + "epoch": 6.836173001310616, + "grad_norm": 1.7314691543579102, + "learning_rate": 1.375992427280503e-05, + "loss": 0.3075, + "num_input_tokens_seen": 6984376, + "step": 26080 + }, + { + "epoch": 6.837483617300131, + "grad_norm": 1.6094599962234497, + "learning_rate": 1.3749709343307401e-05, + "loss": 0.2843, + "num_input_tokens_seen": 6985576, + "step": 26085 + }, + { + "epoch": 6.838794233289646, + "grad_norm": 1.6402642726898193, + "learning_rate": 1.3739496768476168e-05, + "loss": 0.2081, + "num_input_tokens_seen": 6986888, + "step": 26090 + }, + { + "epoch": 6.840104849279161, + "grad_norm": 1.6017743349075317, + "learning_rate": 1.3729286550448814e-05, + "loss": 0.3336, + "num_input_tokens_seen": 6988648, + "step": 26095 + }, + { + "epoch": 6.8414154652686765, + "grad_norm": 1.5667835474014282, + "learning_rate": 1.3719078691362302e-05, + "loss": 0.2324, + "num_input_tokens_seen": 6990264, + "step": 26100 + }, + { + "epoch": 6.842726081258191, + "grad_norm": 0.36744022369384766, + "learning_rate": 1.3708873193353125e-05, + "loss": 0.2177, + "num_input_tokens_seen": 6991256, + "step": 26105 + }, + { + "epoch": 6.844036697247707, + "grad_norm": 1.1623249053955078, + "learning_rate": 1.3698670058557281e-05, + "loss": 0.2325, + "num_input_tokens_seen": 6992360, + "step": 26110 + }, + { + "epoch": 6.845347313237221, + "grad_norm": 0.6173955798149109, + "learning_rate": 1.3688469289110268e-05, + "loss": 0.4432, + "num_input_tokens_seen": 6994360, + "step": 26115 + }, + { + "epoch": 6.846657929226737, + "grad_norm": 1.2765285968780518, + "learning_rate": 1.3678270887147085e-05, + "loss": 0.2254, + "num_input_tokens_seen": 6995464, + "step": 26120 + }, + { + "epoch": 6.8479685452162515, + "grad_norm": 2.3127048015594482, + "learning_rate": 1.3668074854802243e-05, + "loss": 0.2464, + "num_input_tokens_seen": 6996872, + "step": 26125 + }, + { + "epoch": 6.849279161205767, + "grad_norm": 3.770904779434204, + "learning_rate": 1.3657881194209754e-05, + "loss": 0.4621, + "num_input_tokens_seen": 6998248, + "step": 26130 + }, + { + "epoch": 6.850589777195282, + "grad_norm": 3.492521286010742, + "learning_rate": 1.3647689907503136e-05, + "loss": 0.2147, + "num_input_tokens_seen": 6999352, + "step": 26135 + }, + { + "epoch": 6.851900393184797, + "grad_norm": 1.383539080619812, + "learning_rate": 1.3637500996815402e-05, + "loss": 0.2626, + "num_input_tokens_seen": 7000760, + "step": 26140 + }, + { + "epoch": 6.853211009174312, + "grad_norm": 0.8311824798583984, + "learning_rate": 1.3627314464279078e-05, + "loss": 0.2265, + "num_input_tokens_seen": 7002200, + "step": 26145 + }, + { + "epoch": 6.854521625163827, + "grad_norm": 1.9054248332977295, + "learning_rate": 1.3617130312026188e-05, + "loss": 0.3363, + "num_input_tokens_seen": 7004136, + "step": 26150 + }, + { + "epoch": 6.855832241153342, + "grad_norm": 2.018686532974243, + "learning_rate": 1.3606948542188252e-05, + "loss": 0.235, + "num_input_tokens_seen": 7005288, + "step": 26155 + }, + { + "epoch": 6.857142857142857, + "grad_norm": 2.4827189445495605, + "learning_rate": 1.3596769156896303e-05, + "loss": 0.2687, + "num_input_tokens_seen": 7006264, + "step": 26160 + }, + { + "epoch": 6.858453473132372, + "grad_norm": 0.9132869839668274, + "learning_rate": 1.3586592158280864e-05, + "loss": 0.2006, + "num_input_tokens_seen": 7007608, + "step": 26165 + }, + { + "epoch": 6.859764089121887, + "grad_norm": 1.0305185317993164, + "learning_rate": 1.3576417548471967e-05, + "loss": 0.247, + "num_input_tokens_seen": 7009224, + "step": 26170 + }, + { + "epoch": 6.8610747051114025, + "grad_norm": 1.4106639623641968, + "learning_rate": 1.3566245329599148e-05, + "loss": 0.2725, + "num_input_tokens_seen": 7010728, + "step": 26175 + }, + { + "epoch": 6.862385321100917, + "grad_norm": 1.3714165687561035, + "learning_rate": 1.3556075503791416e-05, + "loss": 0.2318, + "num_input_tokens_seen": 7011752, + "step": 26180 + }, + { + "epoch": 6.863695937090433, + "grad_norm": 1.6605579853057861, + "learning_rate": 1.3545908073177308e-05, + "loss": 0.4051, + "num_input_tokens_seen": 7013016, + "step": 26185 + }, + { + "epoch": 6.865006553079947, + "grad_norm": 2.4058241844177246, + "learning_rate": 1.3535743039884841e-05, + "loss": 0.2199, + "num_input_tokens_seen": 7014360, + "step": 26190 + }, + { + "epoch": 6.866317169069463, + "grad_norm": 1.399570345878601, + "learning_rate": 1.3525580406041552e-05, + "loss": 0.3655, + "num_input_tokens_seen": 7015304, + "step": 26195 + }, + { + "epoch": 6.8676277850589775, + "grad_norm": 1.5745152235031128, + "learning_rate": 1.3515420173774463e-05, + "loss": 0.4175, + "num_input_tokens_seen": 7017096, + "step": 26200 + }, + { + "epoch": 6.868938401048493, + "grad_norm": 2.362544059753418, + "learning_rate": 1.3505262345210084e-05, + "loss": 0.834, + "num_input_tokens_seen": 7018248, + "step": 26205 + }, + { + "epoch": 6.870249017038008, + "grad_norm": 2.015044689178467, + "learning_rate": 1.3495106922474444e-05, + "loss": 0.3103, + "num_input_tokens_seen": 7019960, + "step": 26210 + }, + { + "epoch": 6.871559633027523, + "grad_norm": 3.2515146732330322, + "learning_rate": 1.3484953907693038e-05, + "loss": 0.3183, + "num_input_tokens_seen": 7021016, + "step": 26215 + }, + { + "epoch": 6.872870249017038, + "grad_norm": 2.5609636306762695, + "learning_rate": 1.3474803302990885e-05, + "loss": 0.213, + "num_input_tokens_seen": 7022520, + "step": 26220 + }, + { + "epoch": 6.8741808650065535, + "grad_norm": 1.509487509727478, + "learning_rate": 1.3464655110492484e-05, + "loss": 0.2527, + "num_input_tokens_seen": 7024056, + "step": 26225 + }, + { + "epoch": 6.875491480996068, + "grad_norm": 0.6857403516769409, + "learning_rate": 1.3454509332321834e-05, + "loss": 0.1991, + "num_input_tokens_seen": 7025256, + "step": 26230 + }, + { + "epoch": 6.876802096985584, + "grad_norm": 1.213934302330017, + "learning_rate": 1.3444365970602436e-05, + "loss": 0.1787, + "num_input_tokens_seen": 7026424, + "step": 26235 + }, + { + "epoch": 6.878112712975098, + "grad_norm": 1.152869462966919, + "learning_rate": 1.3434225027457274e-05, + "loss": 0.1644, + "num_input_tokens_seen": 7027704, + "step": 26240 + }, + { + "epoch": 6.879423328964613, + "grad_norm": 1.17823326587677, + "learning_rate": 1.3424086505008827e-05, + "loss": 0.1824, + "num_input_tokens_seen": 7029288, + "step": 26245 + }, + { + "epoch": 6.8807339449541285, + "grad_norm": 1.3673601150512695, + "learning_rate": 1.3413950405379075e-05, + "loss": 0.2845, + "num_input_tokens_seen": 7030648, + "step": 26250 + }, + { + "epoch": 6.882044560943643, + "grad_norm": 2.0010805130004883, + "learning_rate": 1.3403816730689483e-05, + "loss": 0.2047, + "num_input_tokens_seen": 7031992, + "step": 26255 + }, + { + "epoch": 6.883355176933159, + "grad_norm": 1.6416940689086914, + "learning_rate": 1.3393685483061014e-05, + "loss": 0.1939, + "num_input_tokens_seen": 7033144, + "step": 26260 + }, + { + "epoch": 6.884665792922673, + "grad_norm": 1.5538403987884521, + "learning_rate": 1.3383556664614117e-05, + "loss": 0.3043, + "num_input_tokens_seen": 7034824, + "step": 26265 + }, + { + "epoch": 6.885976408912189, + "grad_norm": 1.0785831212997437, + "learning_rate": 1.337343027746874e-05, + "loss": 0.2419, + "num_input_tokens_seen": 7036008, + "step": 26270 + }, + { + "epoch": 6.8872870249017035, + "grad_norm": 0.9664319753646851, + "learning_rate": 1.3363306323744315e-05, + "loss": 0.2521, + "num_input_tokens_seen": 7037624, + "step": 26275 + }, + { + "epoch": 6.888597640891219, + "grad_norm": 0.8724886178970337, + "learning_rate": 1.335318480555977e-05, + "loss": 0.3852, + "num_input_tokens_seen": 7039096, + "step": 26280 + }, + { + "epoch": 6.889908256880734, + "grad_norm": 1.180213451385498, + "learning_rate": 1.3343065725033522e-05, + "loss": 0.3295, + "num_input_tokens_seen": 7040664, + "step": 26285 + }, + { + "epoch": 6.891218872870249, + "grad_norm": 1.8864221572875977, + "learning_rate": 1.3332949084283478e-05, + "loss": 0.4118, + "num_input_tokens_seen": 7042136, + "step": 26290 + }, + { + "epoch": 6.892529488859764, + "grad_norm": 2.286379098892212, + "learning_rate": 1.3322834885427038e-05, + "loss": 0.2723, + "num_input_tokens_seen": 7043464, + "step": 26295 + }, + { + "epoch": 6.8938401048492794, + "grad_norm": 3.009868860244751, + "learning_rate": 1.3312723130581073e-05, + "loss": 0.2431, + "num_input_tokens_seen": 7044568, + "step": 26300 + }, + { + "epoch": 6.895150720838794, + "grad_norm": 1.1602656841278076, + "learning_rate": 1.3302613821861951e-05, + "loss": 0.1482, + "num_input_tokens_seen": 7045896, + "step": 26305 + }, + { + "epoch": 6.89646133682831, + "grad_norm": 1.2478128671646118, + "learning_rate": 1.3292506961385553e-05, + "loss": 0.2229, + "num_input_tokens_seen": 7047288, + "step": 26310 + }, + { + "epoch": 6.897771952817824, + "grad_norm": 11.87331771850586, + "learning_rate": 1.3282402551267223e-05, + "loss": 0.3264, + "num_input_tokens_seen": 7048472, + "step": 26315 + }, + { + "epoch": 6.89908256880734, + "grad_norm": 4.026345729827881, + "learning_rate": 1.3272300593621789e-05, + "loss": 0.2255, + "num_input_tokens_seen": 7049832, + "step": 26320 + }, + { + "epoch": 6.9003931847968545, + "grad_norm": 1.1975786685943604, + "learning_rate": 1.3262201090563581e-05, + "loss": 0.3154, + "num_input_tokens_seen": 7051176, + "step": 26325 + }, + { + "epoch": 6.90170380078637, + "grad_norm": 1.1311310529708862, + "learning_rate": 1.325210404420641e-05, + "loss": 0.2527, + "num_input_tokens_seen": 7052888, + "step": 26330 + }, + { + "epoch": 6.903014416775885, + "grad_norm": 1.2431950569152832, + "learning_rate": 1.324200945666355e-05, + "loss": 0.1819, + "num_input_tokens_seen": 7053960, + "step": 26335 + }, + { + "epoch": 6.904325032765399, + "grad_norm": 2.7181808948516846, + "learning_rate": 1.3231917330047802e-05, + "loss": 0.3428, + "num_input_tokens_seen": 7055160, + "step": 26340 + }, + { + "epoch": 6.905635648754915, + "grad_norm": 1.1996119022369385, + "learning_rate": 1.322182766647142e-05, + "loss": 0.2158, + "num_input_tokens_seen": 7056344, + "step": 26345 + }, + { + "epoch": 6.9069462647444295, + "grad_norm": 3.4121880531311035, + "learning_rate": 1.3211740468046158e-05, + "loss": 0.2627, + "num_input_tokens_seen": 7057368, + "step": 26350 + }, + { + "epoch": 6.908256880733945, + "grad_norm": 3.1730141639709473, + "learning_rate": 1.3201655736883244e-05, + "loss": 0.2628, + "num_input_tokens_seen": 7058696, + "step": 26355 + }, + { + "epoch": 6.90956749672346, + "grad_norm": 2.552340269088745, + "learning_rate": 1.31915734750934e-05, + "loss": 0.1913, + "num_input_tokens_seen": 7060216, + "step": 26360 + }, + { + "epoch": 6.910878112712975, + "grad_norm": 1.0850647687911987, + "learning_rate": 1.3181493684786825e-05, + "loss": 0.3625, + "num_input_tokens_seen": 7062168, + "step": 26365 + }, + { + "epoch": 6.91218872870249, + "grad_norm": 1.7869987487792969, + "learning_rate": 1.31714163680732e-05, + "loss": 0.2373, + "num_input_tokens_seen": 7063416, + "step": 26370 + }, + { + "epoch": 6.913499344692005, + "grad_norm": 0.8278384804725647, + "learning_rate": 1.316134152706169e-05, + "loss": 0.1936, + "num_input_tokens_seen": 7064728, + "step": 26375 + }, + { + "epoch": 6.91480996068152, + "grad_norm": 1.8155889511108398, + "learning_rate": 1.3151269163860947e-05, + "loss": 0.4085, + "num_input_tokens_seen": 7067624, + "step": 26380 + }, + { + "epoch": 6.916120576671036, + "grad_norm": 1.6926790475845337, + "learning_rate": 1.3141199280579092e-05, + "loss": 0.5664, + "num_input_tokens_seen": 7068760, + "step": 26385 + }, + { + "epoch": 6.91743119266055, + "grad_norm": 1.7820279598236084, + "learning_rate": 1.3131131879323736e-05, + "loss": 0.2582, + "num_input_tokens_seen": 7070328, + "step": 26390 + }, + { + "epoch": 6.918741808650066, + "grad_norm": 1.9395955801010132, + "learning_rate": 1.3121066962201971e-05, + "loss": 0.1364, + "num_input_tokens_seen": 7071832, + "step": 26395 + }, + { + "epoch": 6.9200524246395805, + "grad_norm": 3.317350149154663, + "learning_rate": 1.3111004531320364e-05, + "loss": 0.1842, + "num_input_tokens_seen": 7072952, + "step": 26400 + }, + { + "epoch": 6.921363040629096, + "grad_norm": 2.189103364944458, + "learning_rate": 1.3100944588784969e-05, + "loss": 0.2755, + "num_input_tokens_seen": 7074488, + "step": 26405 + }, + { + "epoch": 6.922673656618611, + "grad_norm": 1.5467675924301147, + "learning_rate": 1.3090887136701319e-05, + "loss": 0.0885, + "num_input_tokens_seen": 7077080, + "step": 26410 + }, + { + "epoch": 6.923984272608125, + "grad_norm": 1.3700178861618042, + "learning_rate": 1.3080832177174406e-05, + "loss": 0.1999, + "num_input_tokens_seen": 7078616, + "step": 26415 + }, + { + "epoch": 6.925294888597641, + "grad_norm": 4.444522857666016, + "learning_rate": 1.3070779712308711e-05, + "loss": 0.3089, + "num_input_tokens_seen": 7080120, + "step": 26420 + }, + { + "epoch": 6.926605504587156, + "grad_norm": 1.7942675352096558, + "learning_rate": 1.3060729744208222e-05, + "loss": 0.3221, + "num_input_tokens_seen": 7081784, + "step": 26425 + }, + { + "epoch": 6.927916120576671, + "grad_norm": 1.2293496131896973, + "learning_rate": 1.3050682274976369e-05, + "loss": 0.1858, + "num_input_tokens_seen": 7083208, + "step": 26430 + }, + { + "epoch": 6.929226736566186, + "grad_norm": 2.202420711517334, + "learning_rate": 1.304063730671607e-05, + "loss": 0.2216, + "num_input_tokens_seen": 7084456, + "step": 26435 + }, + { + "epoch": 6.930537352555701, + "grad_norm": 3.1053929328918457, + "learning_rate": 1.3030594841529714e-05, + "loss": 0.2964, + "num_input_tokens_seen": 7085800, + "step": 26440 + }, + { + "epoch": 6.931847968545216, + "grad_norm": 1.3606010675430298, + "learning_rate": 1.3020554881519187e-05, + "loss": 0.1826, + "num_input_tokens_seen": 7087144, + "step": 26445 + }, + { + "epoch": 6.933158584534731, + "grad_norm": 0.7551866173744202, + "learning_rate": 1.3010517428785818e-05, + "loss": 0.2094, + "num_input_tokens_seen": 7088632, + "step": 26450 + }, + { + "epoch": 6.934469200524246, + "grad_norm": 1.7008097171783447, + "learning_rate": 1.3000482485430433e-05, + "loss": 0.1839, + "num_input_tokens_seen": 7089800, + "step": 26455 + }, + { + "epoch": 6.935779816513762, + "grad_norm": 1.7142921686172485, + "learning_rate": 1.299045005355333e-05, + "loss": 0.2516, + "num_input_tokens_seen": 7091016, + "step": 26460 + }, + { + "epoch": 6.937090432503276, + "grad_norm": 2.899062395095825, + "learning_rate": 1.2980420135254279e-05, + "loss": 0.3422, + "num_input_tokens_seen": 7092264, + "step": 26465 + }, + { + "epoch": 6.938401048492792, + "grad_norm": 1.1070475578308105, + "learning_rate": 1.2970392732632527e-05, + "loss": 0.2915, + "num_input_tokens_seen": 7093320, + "step": 26470 + }, + { + "epoch": 6.9397116644823065, + "grad_norm": 1.0474703311920166, + "learning_rate": 1.2960367847786781e-05, + "loss": 0.2025, + "num_input_tokens_seen": 7094712, + "step": 26475 + }, + { + "epoch": 6.941022280471822, + "grad_norm": 1.6993420124053955, + "learning_rate": 1.295034548281526e-05, + "loss": 0.231, + "num_input_tokens_seen": 7095880, + "step": 26480 + }, + { + "epoch": 6.942332896461337, + "grad_norm": 1.9141854047775269, + "learning_rate": 1.2940325639815599e-05, + "loss": 0.2552, + "num_input_tokens_seen": 7097288, + "step": 26485 + }, + { + "epoch": 6.943643512450852, + "grad_norm": 2.066490411758423, + "learning_rate": 1.2930308320884946e-05, + "loss": 0.1993, + "num_input_tokens_seen": 7098424, + "step": 26490 + }, + { + "epoch": 6.944954128440367, + "grad_norm": 0.8337986469268799, + "learning_rate": 1.2920293528119903e-05, + "loss": 0.1286, + "num_input_tokens_seen": 7099608, + "step": 26495 + }, + { + "epoch": 6.946264744429882, + "grad_norm": 1.4607923030853271, + "learning_rate": 1.2910281263616559e-05, + "loss": 0.4254, + "num_input_tokens_seen": 7101208, + "step": 26500 + }, + { + "epoch": 6.947575360419397, + "grad_norm": 2.0547053813934326, + "learning_rate": 1.2900271529470451e-05, + "loss": 0.2543, + "num_input_tokens_seen": 7102616, + "step": 26505 + }, + { + "epoch": 6.948885976408912, + "grad_norm": 2.7529730796813965, + "learning_rate": 1.2890264327776608e-05, + "loss": 0.2202, + "num_input_tokens_seen": 7103720, + "step": 26510 + }, + { + "epoch": 6.950196592398427, + "grad_norm": 1.8177945613861084, + "learning_rate": 1.2880259660629516e-05, + "loss": 0.2824, + "num_input_tokens_seen": 7104888, + "step": 26515 + }, + { + "epoch": 6.951507208387943, + "grad_norm": 2.9812891483306885, + "learning_rate": 1.2870257530123136e-05, + "loss": 0.2054, + "num_input_tokens_seen": 7106088, + "step": 26520 + }, + { + "epoch": 6.952817824377457, + "grad_norm": 1.6880247592926025, + "learning_rate": 1.2860257938350895e-05, + "loss": 0.4687, + "num_input_tokens_seen": 7107320, + "step": 26525 + }, + { + "epoch": 6.954128440366972, + "grad_norm": 3.2597498893737793, + "learning_rate": 1.2850260887405702e-05, + "loss": 0.3916, + "num_input_tokens_seen": 7109000, + "step": 26530 + }, + { + "epoch": 6.955439056356488, + "grad_norm": 0.8593471050262451, + "learning_rate": 1.2840266379379896e-05, + "loss": 0.1656, + "num_input_tokens_seen": 7110504, + "step": 26535 + }, + { + "epoch": 6.956749672346002, + "grad_norm": 2.471667528152466, + "learning_rate": 1.2830274416365334e-05, + "loss": 0.303, + "num_input_tokens_seen": 7111912, + "step": 26540 + }, + { + "epoch": 6.958060288335518, + "grad_norm": 1.74921715259552, + "learning_rate": 1.2820285000453309e-05, + "loss": 0.2662, + "num_input_tokens_seen": 7114440, + "step": 26545 + }, + { + "epoch": 6.9593709043250325, + "grad_norm": 2.1702280044555664, + "learning_rate": 1.281029813373459e-05, + "loss": 0.1455, + "num_input_tokens_seen": 7115928, + "step": 26550 + }, + { + "epoch": 6.960681520314548, + "grad_norm": 1.8659064769744873, + "learning_rate": 1.2800313818299403e-05, + "loss": 0.368, + "num_input_tokens_seen": 7117144, + "step": 26555 + }, + { + "epoch": 6.961992136304063, + "grad_norm": 3.090970993041992, + "learning_rate": 1.2790332056237458e-05, + "loss": 0.4449, + "num_input_tokens_seen": 7118552, + "step": 26560 + }, + { + "epoch": 6.963302752293578, + "grad_norm": 2.8406388759613037, + "learning_rate": 1.2780352849637928e-05, + "loss": 0.2255, + "num_input_tokens_seen": 7119640, + "step": 26565 + }, + { + "epoch": 6.964613368283093, + "grad_norm": 1.315902590751648, + "learning_rate": 1.2770376200589423e-05, + "loss": 0.2999, + "num_input_tokens_seen": 7121192, + "step": 26570 + }, + { + "epoch": 6.965923984272608, + "grad_norm": 2.985426902770996, + "learning_rate": 1.2760402111180048e-05, + "loss": 0.4421, + "num_input_tokens_seen": 7122360, + "step": 26575 + }, + { + "epoch": 6.967234600262123, + "grad_norm": 1.144910454750061, + "learning_rate": 1.2750430583497363e-05, + "loss": 0.3872, + "num_input_tokens_seen": 7124104, + "step": 26580 + }, + { + "epoch": 6.968545216251639, + "grad_norm": 3.941012382507324, + "learning_rate": 1.2740461619628392e-05, + "loss": 0.2346, + "num_input_tokens_seen": 7125496, + "step": 26585 + }, + { + "epoch": 6.969855832241153, + "grad_norm": 1.8398183584213257, + "learning_rate": 1.2730495221659607e-05, + "loss": 0.3429, + "num_input_tokens_seen": 7126792, + "step": 26590 + }, + { + "epoch": 6.971166448230669, + "grad_norm": 6.546040058135986, + "learning_rate": 1.2720531391676987e-05, + "loss": 0.289, + "num_input_tokens_seen": 7128536, + "step": 26595 + }, + { + "epoch": 6.972477064220183, + "grad_norm": 4.776291370391846, + "learning_rate": 1.2710570131765937e-05, + "loss": 0.214, + "num_input_tokens_seen": 7129896, + "step": 26600 + }, + { + "epoch": 6.973787680209698, + "grad_norm": 2.7816081047058105, + "learning_rate": 1.2700611444011312e-05, + "loss": 0.2826, + "num_input_tokens_seen": 7131080, + "step": 26605 + }, + { + "epoch": 6.975098296199214, + "grad_norm": 1.1381192207336426, + "learning_rate": 1.2690655330497464e-05, + "loss": 0.201, + "num_input_tokens_seen": 7132504, + "step": 26610 + }, + { + "epoch": 6.976408912188729, + "grad_norm": 2.4970455169677734, + "learning_rate": 1.2680701793308187e-05, + "loss": 0.1505, + "num_input_tokens_seen": 7133720, + "step": 26615 + }, + { + "epoch": 6.977719528178244, + "grad_norm": 2.4909098148345947, + "learning_rate": 1.2670750834526735e-05, + "loss": 0.2833, + "num_input_tokens_seen": 7135176, + "step": 26620 + }, + { + "epoch": 6.9790301441677585, + "grad_norm": 1.9724581241607666, + "learning_rate": 1.266080245623583e-05, + "loss": 0.3211, + "num_input_tokens_seen": 7136648, + "step": 26625 + }, + { + "epoch": 6.980340760157274, + "grad_norm": 1.9480855464935303, + "learning_rate": 1.2650856660517652e-05, + "loss": 0.4187, + "num_input_tokens_seen": 7137640, + "step": 26630 + }, + { + "epoch": 6.981651376146789, + "grad_norm": 2.471569776535034, + "learning_rate": 1.2640913449453837e-05, + "loss": 0.2817, + "num_input_tokens_seen": 7138824, + "step": 26635 + }, + { + "epoch": 6.982961992136304, + "grad_norm": 1.7238268852233887, + "learning_rate": 1.2630972825125482e-05, + "loss": 0.3284, + "num_input_tokens_seen": 7140376, + "step": 26640 + }, + { + "epoch": 6.984272608125819, + "grad_norm": 2.187509059906006, + "learning_rate": 1.2621034789613151e-05, + "loss": 0.4298, + "num_input_tokens_seen": 7141896, + "step": 26645 + }, + { + "epoch": 6.985583224115334, + "grad_norm": 1.5329030752182007, + "learning_rate": 1.2611099344996832e-05, + "loss": 0.2714, + "num_input_tokens_seen": 7143432, + "step": 26650 + }, + { + "epoch": 6.986893840104849, + "grad_norm": 2.1978275775909424, + "learning_rate": 1.2601166493356023e-05, + "loss": 0.2664, + "num_input_tokens_seen": 7144520, + "step": 26655 + }, + { + "epoch": 6.988204456094365, + "grad_norm": 2.094843864440918, + "learning_rate": 1.2591236236769643e-05, + "loss": 0.2432, + "num_input_tokens_seen": 7145816, + "step": 26660 + }, + { + "epoch": 6.989515072083879, + "grad_norm": 1.7676910161972046, + "learning_rate": 1.2581308577316079e-05, + "loss": 0.2753, + "num_input_tokens_seen": 7147192, + "step": 26665 + }, + { + "epoch": 6.990825688073395, + "grad_norm": 2.4682281017303467, + "learning_rate": 1.2571383517073172e-05, + "loss": 0.2579, + "num_input_tokens_seen": 7148488, + "step": 26670 + }, + { + "epoch": 6.992136304062909, + "grad_norm": 1.9382832050323486, + "learning_rate": 1.2561461058118219e-05, + "loss": 0.3096, + "num_input_tokens_seen": 7149256, + "step": 26675 + }, + { + "epoch": 6.993446920052425, + "grad_norm": 3.184429168701172, + "learning_rate": 1.2551541202527984e-05, + "loss": 0.2099, + "num_input_tokens_seen": 7151000, + "step": 26680 + }, + { + "epoch": 6.99475753604194, + "grad_norm": 1.1752926111221313, + "learning_rate": 1.2541623952378657e-05, + "loss": 0.2366, + "num_input_tokens_seen": 7152008, + "step": 26685 + }, + { + "epoch": 6.996068152031455, + "grad_norm": 1.014159083366394, + "learning_rate": 1.2531709309745907e-05, + "loss": 0.197, + "num_input_tokens_seen": 7153320, + "step": 26690 + }, + { + "epoch": 6.99737876802097, + "grad_norm": 3.270343780517578, + "learning_rate": 1.2521797276704858e-05, + "loss": 0.4201, + "num_input_tokens_seen": 7154424, + "step": 26695 + }, + { + "epoch": 6.9986893840104845, + "grad_norm": 1.3583858013153076, + "learning_rate": 1.251188785533008e-05, + "loss": 0.3076, + "num_input_tokens_seen": 7156472, + "step": 26700 + }, + { + "epoch": 7.0, + "grad_norm": 3.1711061000823975, + "learning_rate": 1.2501981047695583e-05, + "loss": 0.2119, + "num_input_tokens_seen": 7157624, + "step": 26705 + }, + { + "epoch": 7.001310615989515, + "grad_norm": 5.384791851043701, + "learning_rate": 1.2492076855874868e-05, + "loss": 0.1318, + "num_input_tokens_seen": 7158776, + "step": 26710 + }, + { + "epoch": 7.001834862385321, + "eval_loss": 0.657537043094635, + "eval_runtime": 16.8719, + "eval_samples_per_second": 50.261, + "eval_steps_per_second": 25.131, + "num_input_tokens_seen": 7159688, + "step": 26712 + }, + { + "epoch": 7.00262123197903, + "grad_norm": 0.9551830291748047, + "learning_rate": 1.2482175281940855e-05, + "loss": 0.1587, + "num_input_tokens_seen": 7160488, + "step": 26715 + }, + { + "epoch": 7.003931847968545, + "grad_norm": 1.5035792589187622, + "learning_rate": 1.2472276327965938e-05, + "loss": 0.2289, + "num_input_tokens_seen": 7162232, + "step": 26720 + }, + { + "epoch": 7.00524246395806, + "grad_norm": 1.6932624578475952, + "learning_rate": 1.2462379996021928e-05, + "loss": 0.2971, + "num_input_tokens_seen": 7164264, + "step": 26725 + }, + { + "epoch": 7.006553079947575, + "grad_norm": 2.832425832748413, + "learning_rate": 1.2452486288180124e-05, + "loss": 0.1329, + "num_input_tokens_seen": 7165720, + "step": 26730 + }, + { + "epoch": 7.007863695937091, + "grad_norm": 1.7647193670272827, + "learning_rate": 1.2442595206511257e-05, + "loss": 0.2393, + "num_input_tokens_seen": 7167048, + "step": 26735 + }, + { + "epoch": 7.009174311926605, + "grad_norm": 1.6048064231872559, + "learning_rate": 1.2432706753085516e-05, + "loss": 0.2148, + "num_input_tokens_seen": 7169224, + "step": 26740 + }, + { + "epoch": 7.010484927916121, + "grad_norm": 1.787567377090454, + "learning_rate": 1.2422820929972537e-05, + "loss": 0.1695, + "num_input_tokens_seen": 7170472, + "step": 26745 + }, + { + "epoch": 7.011795543905635, + "grad_norm": 0.7249892950057983, + "learning_rate": 1.2412937739241407e-05, + "loss": 0.1649, + "num_input_tokens_seen": 7171416, + "step": 26750 + }, + { + "epoch": 7.013106159895151, + "grad_norm": 1.4916176795959473, + "learning_rate": 1.2403057182960655e-05, + "loss": 0.1154, + "num_input_tokens_seen": 7172856, + "step": 26755 + }, + { + "epoch": 7.014416775884666, + "grad_norm": 3.82715106010437, + "learning_rate": 1.2393179263198273e-05, + "loss": 0.1817, + "num_input_tokens_seen": 7174200, + "step": 26760 + }, + { + "epoch": 7.015727391874181, + "grad_norm": 0.9927800893783569, + "learning_rate": 1.2383303982021688e-05, + "loss": 0.1996, + "num_input_tokens_seen": 7175384, + "step": 26765 + }, + { + "epoch": 7.017038007863696, + "grad_norm": 2.1193244457244873, + "learning_rate": 1.237343134149778e-05, + "loss": 0.1808, + "num_input_tokens_seen": 7176824, + "step": 26770 + }, + { + "epoch": 7.018348623853211, + "grad_norm": 1.9744313955307007, + "learning_rate": 1.2363561343692873e-05, + "loss": 0.2574, + "num_input_tokens_seen": 7178040, + "step": 26775 + }, + { + "epoch": 7.019659239842726, + "grad_norm": 1.1272921562194824, + "learning_rate": 1.2353693990672743e-05, + "loss": 0.1585, + "num_input_tokens_seen": 7179224, + "step": 26780 + }, + { + "epoch": 7.0209698558322415, + "grad_norm": 4.560916900634766, + "learning_rate": 1.2343829284502612e-05, + "loss": 0.2016, + "num_input_tokens_seen": 7180280, + "step": 26785 + }, + { + "epoch": 7.022280471821756, + "grad_norm": 2.504159927368164, + "learning_rate": 1.2333967227247143e-05, + "loss": 0.1226, + "num_input_tokens_seen": 7181624, + "step": 26790 + }, + { + "epoch": 7.023591087811272, + "grad_norm": 2.268130302429199, + "learning_rate": 1.2324107820970445e-05, + "loss": 0.3562, + "num_input_tokens_seen": 7182872, + "step": 26795 + }, + { + "epoch": 7.024901703800786, + "grad_norm": 0.8763956427574158, + "learning_rate": 1.231425106773609e-05, + "loss": 0.1563, + "num_input_tokens_seen": 7184120, + "step": 26800 + }, + { + "epoch": 7.026212319790301, + "grad_norm": 1.8566125631332397, + "learning_rate": 1.2304396969607057e-05, + "loss": 0.1518, + "num_input_tokens_seen": 7185000, + "step": 26805 + }, + { + "epoch": 7.027522935779817, + "grad_norm": 2.2979423999786377, + "learning_rate": 1.2294545528645804e-05, + "loss": 0.1833, + "num_input_tokens_seen": 7186456, + "step": 26810 + }, + { + "epoch": 7.028833551769331, + "grad_norm": 1.4442689418792725, + "learning_rate": 1.2284696746914216e-05, + "loss": 0.2847, + "num_input_tokens_seen": 7187496, + "step": 26815 + }, + { + "epoch": 7.030144167758847, + "grad_norm": 2.1127045154571533, + "learning_rate": 1.227485062647362e-05, + "loss": 0.1424, + "num_input_tokens_seen": 7189144, + "step": 26820 + }, + { + "epoch": 7.031454783748361, + "grad_norm": 0.6128252148628235, + "learning_rate": 1.226500716938481e-05, + "loss": 0.0875, + "num_input_tokens_seen": 7190248, + "step": 26825 + }, + { + "epoch": 7.032765399737877, + "grad_norm": 1.5391769409179688, + "learning_rate": 1.2255166377707995e-05, + "loss": 0.1323, + "num_input_tokens_seen": 7191768, + "step": 26830 + }, + { + "epoch": 7.034076015727392, + "grad_norm": 2.8809449672698975, + "learning_rate": 1.2245328253502844e-05, + "loss": 0.2087, + "num_input_tokens_seen": 7193272, + "step": 26835 + }, + { + "epoch": 7.035386631716907, + "grad_norm": 2.249126434326172, + "learning_rate": 1.2235492798828444e-05, + "loss": 0.2001, + "num_input_tokens_seen": 7194792, + "step": 26840 + }, + { + "epoch": 7.036697247706422, + "grad_norm": 2.5480120182037354, + "learning_rate": 1.2225660015743343e-05, + "loss": 0.249, + "num_input_tokens_seen": 7195992, + "step": 26845 + }, + { + "epoch": 7.038007863695937, + "grad_norm": 2.0885367393493652, + "learning_rate": 1.2215829906305531e-05, + "loss": 0.2134, + "num_input_tokens_seen": 7197544, + "step": 26850 + }, + { + "epoch": 7.039318479685452, + "grad_norm": 2.650590658187866, + "learning_rate": 1.2206002472572428e-05, + "loss": 0.2505, + "num_input_tokens_seen": 7198904, + "step": 26855 + }, + { + "epoch": 7.0406290956749675, + "grad_norm": 2.185753107070923, + "learning_rate": 1.2196177716600904e-05, + "loss": 0.1681, + "num_input_tokens_seen": 7200200, + "step": 26860 + }, + { + "epoch": 7.041939711664482, + "grad_norm": 2.1747193336486816, + "learning_rate": 1.2186355640447259e-05, + "loss": 0.2, + "num_input_tokens_seen": 7201576, + "step": 26865 + }, + { + "epoch": 7.043250327653998, + "grad_norm": 1.0764753818511963, + "learning_rate": 1.2176536246167235e-05, + "loss": 0.1063, + "num_input_tokens_seen": 7203272, + "step": 26870 + }, + { + "epoch": 7.044560943643512, + "grad_norm": 1.1736959218978882, + "learning_rate": 1.2166719535816024e-05, + "loss": 0.1306, + "num_input_tokens_seen": 7204504, + "step": 26875 + }, + { + "epoch": 7.045871559633028, + "grad_norm": 0.9519733190536499, + "learning_rate": 1.2156905511448235e-05, + "loss": 0.0983, + "num_input_tokens_seen": 7205544, + "step": 26880 + }, + { + "epoch": 7.047182175622543, + "grad_norm": 3.6112754344940186, + "learning_rate": 1.2147094175117932e-05, + "loss": 0.3037, + "num_input_tokens_seen": 7206760, + "step": 26885 + }, + { + "epoch": 7.048492791612058, + "grad_norm": 3.420994520187378, + "learning_rate": 1.2137285528878611e-05, + "loss": 0.2542, + "num_input_tokens_seen": 7208264, + "step": 26890 + }, + { + "epoch": 7.049803407601573, + "grad_norm": 2.4450559616088867, + "learning_rate": 1.2127479574783205e-05, + "loss": 0.2864, + "num_input_tokens_seen": 7210936, + "step": 26895 + }, + { + "epoch": 7.051114023591087, + "grad_norm": 1.3825271129608154, + "learning_rate": 1.211767631488408e-05, + "loss": 0.1635, + "num_input_tokens_seen": 7212200, + "step": 26900 + }, + { + "epoch": 7.052424639580603, + "grad_norm": 0.5239824056625366, + "learning_rate": 1.2107875751233045e-05, + "loss": 0.2286, + "num_input_tokens_seen": 7213272, + "step": 26905 + }, + { + "epoch": 7.053735255570118, + "grad_norm": 1.3030891418457031, + "learning_rate": 1.2098077885881336e-05, + "loss": 0.1359, + "num_input_tokens_seen": 7214472, + "step": 26910 + }, + { + "epoch": 7.055045871559633, + "grad_norm": 1.527952790260315, + "learning_rate": 1.2088282720879646e-05, + "loss": 0.2132, + "num_input_tokens_seen": 7215608, + "step": 26915 + }, + { + "epoch": 7.056356487549148, + "grad_norm": 2.8131420612335205, + "learning_rate": 1.2078490258278061e-05, + "loss": 0.2241, + "num_input_tokens_seen": 7216952, + "step": 26920 + }, + { + "epoch": 7.057667103538663, + "grad_norm": 2.9543471336364746, + "learning_rate": 1.206870050012614e-05, + "loss": 0.1988, + "num_input_tokens_seen": 7218824, + "step": 26925 + }, + { + "epoch": 7.058977719528178, + "grad_norm": 3.18613600730896, + "learning_rate": 1.205891344847286e-05, + "loss": 0.1666, + "num_input_tokens_seen": 7220136, + "step": 26930 + }, + { + "epoch": 7.0602883355176935, + "grad_norm": 5.861530303955078, + "learning_rate": 1.2049129105366625e-05, + "loss": 0.2273, + "num_input_tokens_seen": 7221112, + "step": 26935 + }, + { + "epoch": 7.061598951507208, + "grad_norm": 0.9379897713661194, + "learning_rate": 1.20393474728553e-05, + "loss": 0.1948, + "num_input_tokens_seen": 7222696, + "step": 26940 + }, + { + "epoch": 7.062909567496724, + "grad_norm": 1.206250548362732, + "learning_rate": 1.2029568552986156e-05, + "loss": 0.1658, + "num_input_tokens_seen": 7223928, + "step": 26945 + }, + { + "epoch": 7.064220183486238, + "grad_norm": 4.548574447631836, + "learning_rate": 1.2019792347805902e-05, + "loss": 0.1991, + "num_input_tokens_seen": 7225160, + "step": 26950 + }, + { + "epoch": 7.065530799475754, + "grad_norm": 1.4746313095092773, + "learning_rate": 1.2010018859360691e-05, + "loss": 0.2895, + "num_input_tokens_seen": 7226728, + "step": 26955 + }, + { + "epoch": 7.066841415465269, + "grad_norm": 1.5060397386550903, + "learning_rate": 1.2000248089696078e-05, + "loss": 0.1613, + "num_input_tokens_seen": 7227944, + "step": 26960 + }, + { + "epoch": 7.068152031454784, + "grad_norm": 1.7275047302246094, + "learning_rate": 1.1990480040857076e-05, + "loss": 0.1851, + "num_input_tokens_seen": 7229208, + "step": 26965 + }, + { + "epoch": 7.069462647444299, + "grad_norm": 1.373394250869751, + "learning_rate": 1.1980714714888127e-05, + "loss": 0.1207, + "num_input_tokens_seen": 7230616, + "step": 26970 + }, + { + "epoch": 7.070773263433814, + "grad_norm": 2.6705141067504883, + "learning_rate": 1.197095211383309e-05, + "loss": 0.2517, + "num_input_tokens_seen": 7231896, + "step": 26975 + }, + { + "epoch": 7.072083879423329, + "grad_norm": 1.480402946472168, + "learning_rate": 1.1961192239735264e-05, + "loss": 0.3097, + "num_input_tokens_seen": 7233496, + "step": 26980 + }, + { + "epoch": 7.073394495412844, + "grad_norm": 1.3364379405975342, + "learning_rate": 1.1951435094637373e-05, + "loss": 0.1559, + "num_input_tokens_seen": 7234936, + "step": 26985 + }, + { + "epoch": 7.074705111402359, + "grad_norm": 1.8131951093673706, + "learning_rate": 1.194168068058157e-05, + "loss": 0.1997, + "num_input_tokens_seen": 7236200, + "step": 26990 + }, + { + "epoch": 7.076015727391874, + "grad_norm": 1.5285817384719849, + "learning_rate": 1.1931928999609438e-05, + "loss": 0.1308, + "num_input_tokens_seen": 7237672, + "step": 26995 + }, + { + "epoch": 7.077326343381389, + "grad_norm": 0.9286881685256958, + "learning_rate": 1.1922180053761985e-05, + "loss": 0.1822, + "num_input_tokens_seen": 7238952, + "step": 27000 + }, + { + "epoch": 7.078636959370904, + "grad_norm": 1.929876446723938, + "learning_rate": 1.1912433845079652e-05, + "loss": 0.3459, + "num_input_tokens_seen": 7240296, + "step": 27005 + }, + { + "epoch": 7.0799475753604195, + "grad_norm": 2.5999093055725098, + "learning_rate": 1.19026903756023e-05, + "loss": 0.2009, + "num_input_tokens_seen": 7241448, + "step": 27010 + }, + { + "epoch": 7.081258191349934, + "grad_norm": 2.3248417377471924, + "learning_rate": 1.1892949647369223e-05, + "loss": 0.1906, + "num_input_tokens_seen": 7242920, + "step": 27015 + }, + { + "epoch": 7.08256880733945, + "grad_norm": 3.390570878982544, + "learning_rate": 1.1883211662419137e-05, + "loss": 0.1397, + "num_input_tokens_seen": 7244488, + "step": 27020 + }, + { + "epoch": 7.083879423328964, + "grad_norm": 2.0873663425445557, + "learning_rate": 1.1873476422790187e-05, + "loss": 0.2193, + "num_input_tokens_seen": 7245928, + "step": 27025 + }, + { + "epoch": 7.08519003931848, + "grad_norm": 1.8986254930496216, + "learning_rate": 1.1863743930519939e-05, + "loss": 0.269, + "num_input_tokens_seen": 7247208, + "step": 27030 + }, + { + "epoch": 7.086500655307995, + "grad_norm": 6.067636013031006, + "learning_rate": 1.1854014187645398e-05, + "loss": 0.2256, + "num_input_tokens_seen": 7248360, + "step": 27035 + }, + { + "epoch": 7.08781127129751, + "grad_norm": 1.2428520917892456, + "learning_rate": 1.1844287196202967e-05, + "loss": 0.152, + "num_input_tokens_seen": 7249800, + "step": 27040 + }, + { + "epoch": 7.089121887287025, + "grad_norm": 2.296312093734741, + "learning_rate": 1.183456295822849e-05, + "loss": 0.1834, + "num_input_tokens_seen": 7251448, + "step": 27045 + }, + { + "epoch": 7.09043250327654, + "grad_norm": 2.7050278186798096, + "learning_rate": 1.1824841475757232e-05, + "loss": 0.3464, + "num_input_tokens_seen": 7252712, + "step": 27050 + }, + { + "epoch": 7.091743119266055, + "grad_norm": 2.8279552459716797, + "learning_rate": 1.1815122750823896e-05, + "loss": 0.285, + "num_input_tokens_seen": 7254968, + "step": 27055 + }, + { + "epoch": 7.0930537352555705, + "grad_norm": 1.0496290922164917, + "learning_rate": 1.1805406785462587e-05, + "loss": 0.1473, + "num_input_tokens_seen": 7256312, + "step": 27060 + }, + { + "epoch": 7.094364351245085, + "grad_norm": 2.399777889251709, + "learning_rate": 1.1795693581706838e-05, + "loss": 0.1764, + "num_input_tokens_seen": 7257528, + "step": 27065 + }, + { + "epoch": 7.095674967234601, + "grad_norm": 1.7468845844268799, + "learning_rate": 1.1785983141589618e-05, + "loss": 0.1618, + "num_input_tokens_seen": 7258808, + "step": 27070 + }, + { + "epoch": 7.096985583224115, + "grad_norm": 2.0261311531066895, + "learning_rate": 1.1776275467143283e-05, + "loss": 0.2039, + "num_input_tokens_seen": 7260200, + "step": 27075 + }, + { + "epoch": 7.09829619921363, + "grad_norm": 6.465932369232178, + "learning_rate": 1.1766570560399647e-05, + "loss": 0.1725, + "num_input_tokens_seen": 7261320, + "step": 27080 + }, + { + "epoch": 7.0996068152031455, + "grad_norm": 4.636574745178223, + "learning_rate": 1.1756868423389925e-05, + "loss": 0.2187, + "num_input_tokens_seen": 7262696, + "step": 27085 + }, + { + "epoch": 7.10091743119266, + "grad_norm": 2.15800404548645, + "learning_rate": 1.1747169058144761e-05, + "loss": 0.2456, + "num_input_tokens_seen": 7263944, + "step": 27090 + }, + { + "epoch": 7.102228047182176, + "grad_norm": 1.9941012859344482, + "learning_rate": 1.1737472466694213e-05, + "loss": 0.2765, + "num_input_tokens_seen": 7265112, + "step": 27095 + }, + { + "epoch": 7.10353866317169, + "grad_norm": 3.302748918533325, + "learning_rate": 1.1727778651067761e-05, + "loss": 0.13, + "num_input_tokens_seen": 7266424, + "step": 27100 + }, + { + "epoch": 7.104849279161206, + "grad_norm": 1.9196895360946655, + "learning_rate": 1.1718087613294307e-05, + "loss": 0.2217, + "num_input_tokens_seen": 7267848, + "step": 27105 + }, + { + "epoch": 7.1061598951507206, + "grad_norm": 1.1353156566619873, + "learning_rate": 1.1708399355402163e-05, + "loss": 0.1891, + "num_input_tokens_seen": 7269416, + "step": 27110 + }, + { + "epoch": 7.107470511140236, + "grad_norm": 1.5115431547164917, + "learning_rate": 1.1698713879419072e-05, + "loss": 0.2248, + "num_input_tokens_seen": 7271032, + "step": 27115 + }, + { + "epoch": 7.108781127129751, + "grad_norm": 2.771868944168091, + "learning_rate": 1.168903118737218e-05, + "loss": 0.2056, + "num_input_tokens_seen": 7272216, + "step": 27120 + }, + { + "epoch": 7.110091743119266, + "grad_norm": 0.9774207472801208, + "learning_rate": 1.1679351281288062e-05, + "loss": 0.2296, + "num_input_tokens_seen": 7273688, + "step": 27125 + }, + { + "epoch": 7.111402359108781, + "grad_norm": 2.0920510292053223, + "learning_rate": 1.1669674163192703e-05, + "loss": 0.3088, + "num_input_tokens_seen": 7274808, + "step": 27130 + }, + { + "epoch": 7.1127129750982965, + "grad_norm": 4.238604545593262, + "learning_rate": 1.165999983511151e-05, + "loss": 0.2584, + "num_input_tokens_seen": 7276664, + "step": 27135 + }, + { + "epoch": 7.114023591087811, + "grad_norm": 1.8468986749649048, + "learning_rate": 1.1650328299069301e-05, + "loss": 0.1587, + "num_input_tokens_seen": 7277912, + "step": 27140 + }, + { + "epoch": 7.115334207077327, + "grad_norm": 2.745957851409912, + "learning_rate": 1.1640659557090313e-05, + "loss": 0.1902, + "num_input_tokens_seen": 7278920, + "step": 27145 + }, + { + "epoch": 7.116644823066841, + "grad_norm": 5.115386486053467, + "learning_rate": 1.1630993611198207e-05, + "loss": 0.2516, + "num_input_tokens_seen": 7280184, + "step": 27150 + }, + { + "epoch": 7.117955439056357, + "grad_norm": 1.311886191368103, + "learning_rate": 1.1621330463416027e-05, + "loss": 0.2499, + "num_input_tokens_seen": 7281736, + "step": 27155 + }, + { + "epoch": 7.1192660550458715, + "grad_norm": 2.2402756214141846, + "learning_rate": 1.1611670115766265e-05, + "loss": 0.1696, + "num_input_tokens_seen": 7283112, + "step": 27160 + }, + { + "epoch": 7.120576671035387, + "grad_norm": 6.994268417358398, + "learning_rate": 1.1602012570270806e-05, + "loss": 0.1217, + "num_input_tokens_seen": 7284136, + "step": 27165 + }, + { + "epoch": 7.121887287024902, + "grad_norm": 1.3624874353408813, + "learning_rate": 1.1592357828950976e-05, + "loss": 0.309, + "num_input_tokens_seen": 7285576, + "step": 27170 + }, + { + "epoch": 7.123197903014416, + "grad_norm": 3.4273674488067627, + "learning_rate": 1.1582705893827483e-05, + "loss": 0.1087, + "num_input_tokens_seen": 7286856, + "step": 27175 + }, + { + "epoch": 7.124508519003932, + "grad_norm": 1.8116110563278198, + "learning_rate": 1.1573056766920465e-05, + "loss": 0.1441, + "num_input_tokens_seen": 7288168, + "step": 27180 + }, + { + "epoch": 7.1258191349934465, + "grad_norm": 1.7439148426055908, + "learning_rate": 1.1563410450249468e-05, + "loss": 0.1315, + "num_input_tokens_seen": 7289448, + "step": 27185 + }, + { + "epoch": 7.127129750982962, + "grad_norm": 3.3648276329040527, + "learning_rate": 1.1553766945833452e-05, + "loss": 0.2045, + "num_input_tokens_seen": 7290744, + "step": 27190 + }, + { + "epoch": 7.128440366972477, + "grad_norm": 1.7013566493988037, + "learning_rate": 1.1544126255690774e-05, + "loss": 0.1627, + "num_input_tokens_seen": 7292296, + "step": 27195 + }, + { + "epoch": 7.129750982961992, + "grad_norm": 1.8426647186279297, + "learning_rate": 1.153448838183922e-05, + "loss": 0.1334, + "num_input_tokens_seen": 7293336, + "step": 27200 + }, + { + "epoch": 7.131061598951507, + "grad_norm": 1.5336534976959229, + "learning_rate": 1.1524853326295982e-05, + "loss": 0.1313, + "num_input_tokens_seen": 7294392, + "step": 27205 + }, + { + "epoch": 7.1323722149410225, + "grad_norm": 0.5396798253059387, + "learning_rate": 1.151522109107766e-05, + "loss": 0.1533, + "num_input_tokens_seen": 7295816, + "step": 27210 + }, + { + "epoch": 7.133682830930537, + "grad_norm": 1.7946642637252808, + "learning_rate": 1.150559167820026e-05, + "loss": 0.3287, + "num_input_tokens_seen": 7297320, + "step": 27215 + }, + { + "epoch": 7.134993446920053, + "grad_norm": 2.168889045715332, + "learning_rate": 1.14959650896792e-05, + "loss": 0.165, + "num_input_tokens_seen": 7298440, + "step": 27220 + }, + { + "epoch": 7.136304062909567, + "grad_norm": 1.2041419744491577, + "learning_rate": 1.1486341327529332e-05, + "loss": 0.1604, + "num_input_tokens_seen": 7300104, + "step": 27225 + }, + { + "epoch": 7.137614678899083, + "grad_norm": 1.7644981145858765, + "learning_rate": 1.1476720393764864e-05, + "loss": 0.1601, + "num_input_tokens_seen": 7301224, + "step": 27230 + }, + { + "epoch": 7.1389252948885975, + "grad_norm": 0.6959578990936279, + "learning_rate": 1.146710229039945e-05, + "loss": 0.1689, + "num_input_tokens_seen": 7302616, + "step": 27235 + }, + { + "epoch": 7.140235910878113, + "grad_norm": 1.7222398519515991, + "learning_rate": 1.1457487019446145e-05, + "loss": 0.2531, + "num_input_tokens_seen": 7303752, + "step": 27240 + }, + { + "epoch": 7.141546526867628, + "grad_norm": 0.8069484233856201, + "learning_rate": 1.1447874582917407e-05, + "loss": 0.1603, + "num_input_tokens_seen": 7304968, + "step": 27245 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 1.7319432497024536, + "learning_rate": 1.1438264982825104e-05, + "loss": 0.1795, + "num_input_tokens_seen": 7306248, + "step": 27250 + }, + { + "epoch": 7.144167758846658, + "grad_norm": 2.976649761199951, + "learning_rate": 1.1428658221180506e-05, + "loss": 0.1637, + "num_input_tokens_seen": 7308120, + "step": 27255 + }, + { + "epoch": 7.145478374836173, + "grad_norm": 0.19746637344360352, + "learning_rate": 1.1419054299994292e-05, + "loss": 0.1886, + "num_input_tokens_seen": 7309144, + "step": 27260 + }, + { + "epoch": 7.146788990825688, + "grad_norm": 2.2985169887542725, + "learning_rate": 1.1409453221276545e-05, + "loss": 0.2679, + "num_input_tokens_seen": 7310632, + "step": 27265 + }, + { + "epoch": 7.148099606815203, + "grad_norm": 1.699202060699463, + "learning_rate": 1.1399854987036768e-05, + "loss": 0.191, + "num_input_tokens_seen": 7311848, + "step": 27270 + }, + { + "epoch": 7.149410222804718, + "grad_norm": 2.5240578651428223, + "learning_rate": 1.1390259599283835e-05, + "loss": 0.2328, + "num_input_tokens_seen": 7313048, + "step": 27275 + }, + { + "epoch": 7.150720838794233, + "grad_norm": 3.699244737625122, + "learning_rate": 1.1380667060026041e-05, + "loss": 0.2022, + "num_input_tokens_seen": 7313992, + "step": 27280 + }, + { + "epoch": 7.1520314547837485, + "grad_norm": 1.9430252313613892, + "learning_rate": 1.1371077371271108e-05, + "loss": 0.1891, + "num_input_tokens_seen": 7315464, + "step": 27285 + }, + { + "epoch": 7.153342070773263, + "grad_norm": 1.5824310779571533, + "learning_rate": 1.136149053502613e-05, + "loss": 0.24, + "num_input_tokens_seen": 7316840, + "step": 27290 + }, + { + "epoch": 7.154652686762779, + "grad_norm": 0.999780535697937, + "learning_rate": 1.1351906553297622e-05, + "loss": 0.2042, + "num_input_tokens_seen": 7318232, + "step": 27295 + }, + { + "epoch": 7.155963302752293, + "grad_norm": 4.960216045379639, + "learning_rate": 1.1342325428091488e-05, + "loss": 0.1329, + "num_input_tokens_seen": 7319320, + "step": 27300 + }, + { + "epoch": 7.157273918741809, + "grad_norm": 2.959812879562378, + "learning_rate": 1.1332747161413054e-05, + "loss": 0.1377, + "num_input_tokens_seen": 7320600, + "step": 27305 + }, + { + "epoch": 7.1585845347313235, + "grad_norm": 1.6008533239364624, + "learning_rate": 1.1323171755267015e-05, + "loss": 0.1657, + "num_input_tokens_seen": 7322072, + "step": 27310 + }, + { + "epoch": 7.159895150720839, + "grad_norm": 2.1082897186279297, + "learning_rate": 1.1313599211657494e-05, + "loss": 0.175, + "num_input_tokens_seen": 7323032, + "step": 27315 + }, + { + "epoch": 7.161205766710354, + "grad_norm": 1.117956519126892, + "learning_rate": 1.130402953258801e-05, + "loss": 0.23, + "num_input_tokens_seen": 7324472, + "step": 27320 + }, + { + "epoch": 7.162516382699869, + "grad_norm": 3.1409432888031006, + "learning_rate": 1.129446272006148e-05, + "loss": 0.3354, + "num_input_tokens_seen": 7325704, + "step": 27325 + }, + { + "epoch": 7.163826998689384, + "grad_norm": 2.102963924407959, + "learning_rate": 1.1284898776080222e-05, + "loss": 0.2089, + "num_input_tokens_seen": 7327368, + "step": 27330 + }, + { + "epoch": 7.165137614678899, + "grad_norm": 5.190706253051758, + "learning_rate": 1.127533770264594e-05, + "loss": 0.2753, + "num_input_tokens_seen": 7328664, + "step": 27335 + }, + { + "epoch": 7.166448230668414, + "grad_norm": 1.6157362461090088, + "learning_rate": 1.1265779501759786e-05, + "loss": 0.2379, + "num_input_tokens_seen": 7329912, + "step": 27340 + }, + { + "epoch": 7.16775884665793, + "grad_norm": 1.7506401538848877, + "learning_rate": 1.125622417542224e-05, + "loss": 0.2183, + "num_input_tokens_seen": 7331208, + "step": 27345 + }, + { + "epoch": 7.169069462647444, + "grad_norm": 1.5951588153839111, + "learning_rate": 1.1246671725633227e-05, + "loss": 0.1394, + "num_input_tokens_seen": 7332216, + "step": 27350 + }, + { + "epoch": 7.17038007863696, + "grad_norm": 0.7443305253982544, + "learning_rate": 1.1237122154392058e-05, + "loss": 0.0815, + "num_input_tokens_seen": 7333784, + "step": 27355 + }, + { + "epoch": 7.1716906946264745, + "grad_norm": 3.087414264678955, + "learning_rate": 1.122757546369744e-05, + "loss": 0.2731, + "num_input_tokens_seen": 7335304, + "step": 27360 + }, + { + "epoch": 7.173001310615989, + "grad_norm": 1.566495656967163, + "learning_rate": 1.1218031655547482e-05, + "loss": 0.1741, + "num_input_tokens_seen": 7336680, + "step": 27365 + }, + { + "epoch": 7.174311926605505, + "grad_norm": 1.374976396560669, + "learning_rate": 1.1208490731939683e-05, + "loss": 0.2598, + "num_input_tokens_seen": 7338088, + "step": 27370 + }, + { + "epoch": 7.175622542595019, + "grad_norm": 1.966117262840271, + "learning_rate": 1.1198952694870945e-05, + "loss": 0.2378, + "num_input_tokens_seen": 7339384, + "step": 27375 + }, + { + "epoch": 7.176933158584535, + "grad_norm": 6.964699745178223, + "learning_rate": 1.1189417546337564e-05, + "loss": 0.296, + "num_input_tokens_seen": 7340648, + "step": 27380 + }, + { + "epoch": 7.1782437745740495, + "grad_norm": 2.123140811920166, + "learning_rate": 1.1179885288335235e-05, + "loss": 0.1499, + "num_input_tokens_seen": 7341912, + "step": 27385 + }, + { + "epoch": 7.179554390563565, + "grad_norm": 3.5976085662841797, + "learning_rate": 1.1170355922859024e-05, + "loss": 0.2171, + "num_input_tokens_seen": 7343272, + "step": 27390 + }, + { + "epoch": 7.18086500655308, + "grad_norm": 1.537083625793457, + "learning_rate": 1.1160829451903418e-05, + "loss": 0.2155, + "num_input_tokens_seen": 7344696, + "step": 27395 + }, + { + "epoch": 7.182175622542595, + "grad_norm": 3.2701637744903564, + "learning_rate": 1.1151305877462304e-05, + "loss": 0.2028, + "num_input_tokens_seen": 7346008, + "step": 27400 + }, + { + "epoch": 7.18348623853211, + "grad_norm": 1.6455715894699097, + "learning_rate": 1.1141785201528936e-05, + "loss": 0.1748, + "num_input_tokens_seen": 7347512, + "step": 27405 + }, + { + "epoch": 7.184796854521625, + "grad_norm": 0.7972441911697388, + "learning_rate": 1.1132267426095985e-05, + "loss": 0.1957, + "num_input_tokens_seen": 7348936, + "step": 27410 + }, + { + "epoch": 7.18610747051114, + "grad_norm": 2.419604539871216, + "learning_rate": 1.11227525531555e-05, + "loss": 0.1258, + "num_input_tokens_seen": 7350184, + "step": 27415 + }, + { + "epoch": 7.187418086500656, + "grad_norm": 1.9590506553649902, + "learning_rate": 1.1113240584698927e-05, + "loss": 0.1542, + "num_input_tokens_seen": 7351432, + "step": 27420 + }, + { + "epoch": 7.18872870249017, + "grad_norm": 1.2943440675735474, + "learning_rate": 1.1103731522717115e-05, + "loss": 0.1408, + "num_input_tokens_seen": 7352504, + "step": 27425 + }, + { + "epoch": 7.190039318479686, + "grad_norm": 1.4907691478729248, + "learning_rate": 1.1094225369200272e-05, + "loss": 0.1875, + "num_input_tokens_seen": 7354024, + "step": 27430 + }, + { + "epoch": 7.1913499344692005, + "grad_norm": 2.078641891479492, + "learning_rate": 1.1084722126138037e-05, + "loss": 0.2945, + "num_input_tokens_seen": 7355336, + "step": 27435 + }, + { + "epoch": 7.192660550458716, + "grad_norm": 1.4633392095565796, + "learning_rate": 1.1075221795519413e-05, + "loss": 0.2579, + "num_input_tokens_seen": 7356696, + "step": 27440 + }, + { + "epoch": 7.193971166448231, + "grad_norm": 1.239060878753662, + "learning_rate": 1.106572437933281e-05, + "loss": 0.1969, + "num_input_tokens_seen": 7358312, + "step": 27445 + }, + { + "epoch": 7.195281782437746, + "grad_norm": 3.2085232734680176, + "learning_rate": 1.1056229879566004e-05, + "loss": 0.1944, + "num_input_tokens_seen": 7359512, + "step": 27450 + }, + { + "epoch": 7.196592398427261, + "grad_norm": 0.9969280362129211, + "learning_rate": 1.1046738298206205e-05, + "loss": 0.2503, + "num_input_tokens_seen": 7360808, + "step": 27455 + }, + { + "epoch": 7.1979030144167755, + "grad_norm": 1.9439753293991089, + "learning_rate": 1.103724963723998e-05, + "loss": 0.2688, + "num_input_tokens_seen": 7362408, + "step": 27460 + }, + { + "epoch": 7.199213630406291, + "grad_norm": 1.9076646566390991, + "learning_rate": 1.1027763898653268e-05, + "loss": 0.1972, + "num_input_tokens_seen": 7364184, + "step": 27465 + }, + { + "epoch": 7.200524246395806, + "grad_norm": 1.0137286186218262, + "learning_rate": 1.1018281084431432e-05, + "loss": 0.1447, + "num_input_tokens_seen": 7365208, + "step": 27470 + }, + { + "epoch": 7.201834862385321, + "grad_norm": 3.342324733734131, + "learning_rate": 1.1008801196559204e-05, + "loss": 0.1481, + "num_input_tokens_seen": 7366264, + "step": 27475 + }, + { + "epoch": 7.203145478374836, + "grad_norm": 6.8167595863342285, + "learning_rate": 1.0999324237020714e-05, + "loss": 0.2973, + "num_input_tokens_seen": 7368328, + "step": 27480 + }, + { + "epoch": 7.204456094364351, + "grad_norm": 3.0711445808410645, + "learning_rate": 1.0989850207799469e-05, + "loss": 0.2165, + "num_input_tokens_seen": 7370296, + "step": 27485 + }, + { + "epoch": 7.205766710353866, + "grad_norm": 5.895941257476807, + "learning_rate": 1.0980379110878366e-05, + "loss": 0.2562, + "num_input_tokens_seen": 7371176, + "step": 27490 + }, + { + "epoch": 7.207077326343382, + "grad_norm": 2.8466312885284424, + "learning_rate": 1.0970910948239694e-05, + "loss": 0.2192, + "num_input_tokens_seen": 7372312, + "step": 27495 + }, + { + "epoch": 7.208387942332896, + "grad_norm": 1.3526400327682495, + "learning_rate": 1.0961445721865116e-05, + "loss": 0.2158, + "num_input_tokens_seen": 7373544, + "step": 27500 + }, + { + "epoch": 7.209698558322412, + "grad_norm": 1.47684645652771, + "learning_rate": 1.0951983433735704e-05, + "loss": 0.1245, + "num_input_tokens_seen": 7374888, + "step": 27505 + }, + { + "epoch": 7.2110091743119265, + "grad_norm": 2.4940097332000732, + "learning_rate": 1.0942524085831868e-05, + "loss": 0.2076, + "num_input_tokens_seen": 7376504, + "step": 27510 + }, + { + "epoch": 7.212319790301442, + "grad_norm": 2.2432878017425537, + "learning_rate": 1.0933067680133457e-05, + "loss": 0.1208, + "num_input_tokens_seen": 7377752, + "step": 27515 + }, + { + "epoch": 7.213630406290957, + "grad_norm": 1.223071813583374, + "learning_rate": 1.092361421861968e-05, + "loss": 0.0839, + "num_input_tokens_seen": 7378952, + "step": 27520 + }, + { + "epoch": 7.214941022280472, + "grad_norm": 1.5130184888839722, + "learning_rate": 1.091416370326912e-05, + "loss": 0.1615, + "num_input_tokens_seen": 7380520, + "step": 27525 + }, + { + "epoch": 7.216251638269987, + "grad_norm": 2.0841314792633057, + "learning_rate": 1.090471613605976e-05, + "loss": 0.2517, + "num_input_tokens_seen": 7382008, + "step": 27530 + }, + { + "epoch": 7.217562254259502, + "grad_norm": 5.5293097496032715, + "learning_rate": 1.0895271518968962e-05, + "loss": 0.1053, + "num_input_tokens_seen": 7383592, + "step": 27535 + }, + { + "epoch": 7.218872870249017, + "grad_norm": 1.483713150024414, + "learning_rate": 1.0885829853973467e-05, + "loss": 0.1452, + "num_input_tokens_seen": 7384728, + "step": 27540 + }, + { + "epoch": 7.220183486238533, + "grad_norm": 1.8339405059814453, + "learning_rate": 1.087639114304939e-05, + "loss": 0.2641, + "num_input_tokens_seen": 7385880, + "step": 27545 + }, + { + "epoch": 7.221494102228047, + "grad_norm": 2.2229461669921875, + "learning_rate": 1.0866955388172245e-05, + "loss": 0.1869, + "num_input_tokens_seen": 7387000, + "step": 27550 + }, + { + "epoch": 7.222804718217562, + "grad_norm": 1.160585641860962, + "learning_rate": 1.0857522591316915e-05, + "loss": 0.1027, + "num_input_tokens_seen": 7388520, + "step": 27555 + }, + { + "epoch": 7.224115334207077, + "grad_norm": 1.3681843280792236, + "learning_rate": 1.084809275445767e-05, + "loss": 0.1551, + "num_input_tokens_seen": 7389832, + "step": 27560 + }, + { + "epoch": 7.225425950196592, + "grad_norm": 3.530259132385254, + "learning_rate": 1.083866587956815e-05, + "loss": 0.2918, + "num_input_tokens_seen": 7390712, + "step": 27565 + }, + { + "epoch": 7.226736566186108, + "grad_norm": 4.054325103759766, + "learning_rate": 1.0829241968621401e-05, + "loss": 0.2457, + "num_input_tokens_seen": 7392344, + "step": 27570 + }, + { + "epoch": 7.228047182175622, + "grad_norm": 1.3647921085357666, + "learning_rate": 1.0819821023589832e-05, + "loss": 0.1967, + "num_input_tokens_seen": 7393736, + "step": 27575 + }, + { + "epoch": 7.229357798165138, + "grad_norm": 3.093308210372925, + "learning_rate": 1.0810403046445214e-05, + "loss": 0.1861, + "num_input_tokens_seen": 7395256, + "step": 27580 + }, + { + "epoch": 7.2306684141546524, + "grad_norm": 0.6968086957931519, + "learning_rate": 1.0800988039158719e-05, + "loss": 0.0758, + "num_input_tokens_seen": 7396712, + "step": 27585 + }, + { + "epoch": 7.231979030144168, + "grad_norm": 2.020190954208374, + "learning_rate": 1.0791576003700895e-05, + "loss": 0.1967, + "num_input_tokens_seen": 7397832, + "step": 27590 + }, + { + "epoch": 7.233289646133683, + "grad_norm": 1.453521728515625, + "learning_rate": 1.0782166942041663e-05, + "loss": 0.1306, + "num_input_tokens_seen": 7398904, + "step": 27595 + }, + { + "epoch": 7.234600262123198, + "grad_norm": 2.1563096046447754, + "learning_rate": 1.0772760856150322e-05, + "loss": 0.2022, + "num_input_tokens_seen": 7399896, + "step": 27600 + }, + { + "epoch": 7.235910878112713, + "grad_norm": 1.8982982635498047, + "learning_rate": 1.0763357747995551e-05, + "loss": 0.2159, + "num_input_tokens_seen": 7401128, + "step": 27605 + }, + { + "epoch": 7.237221494102228, + "grad_norm": 1.5426554679870605, + "learning_rate": 1.0753957619545405e-05, + "loss": 0.3994, + "num_input_tokens_seen": 7402632, + "step": 27610 + }, + { + "epoch": 7.238532110091743, + "grad_norm": 1.6714879274368286, + "learning_rate": 1.0744560472767315e-05, + "loss": 0.165, + "num_input_tokens_seen": 7403960, + "step": 27615 + }, + { + "epoch": 7.239842726081259, + "grad_norm": 2.0796258449554443, + "learning_rate": 1.0735166309628083e-05, + "loss": 0.1357, + "num_input_tokens_seen": 7405064, + "step": 27620 + }, + { + "epoch": 7.241153342070773, + "grad_norm": 1.3666824102401733, + "learning_rate": 1.0725775132093899e-05, + "loss": 0.2458, + "num_input_tokens_seen": 7406504, + "step": 27625 + }, + { + "epoch": 7.242463958060289, + "grad_norm": 1.4632794857025146, + "learning_rate": 1.0716386942130313e-05, + "loss": 0.1627, + "num_input_tokens_seen": 7407928, + "step": 27630 + }, + { + "epoch": 7.243774574049803, + "grad_norm": 1.4109872579574585, + "learning_rate": 1.0707001741702261e-05, + "loss": 0.2083, + "num_input_tokens_seen": 7409432, + "step": 27635 + }, + { + "epoch": 7.245085190039318, + "grad_norm": 2.3990132808685303, + "learning_rate": 1.0697619532774047e-05, + "loss": 0.2418, + "num_input_tokens_seen": 7410600, + "step": 27640 + }, + { + "epoch": 7.246395806028834, + "grad_norm": 1.2708420753479004, + "learning_rate": 1.0688240317309355e-05, + "loss": 0.2335, + "num_input_tokens_seen": 7411736, + "step": 27645 + }, + { + "epoch": 7.247706422018348, + "grad_norm": 2.0765161514282227, + "learning_rate": 1.0678864097271234e-05, + "loss": 0.3001, + "num_input_tokens_seen": 7413304, + "step": 27650 + }, + { + "epoch": 7.249017038007864, + "grad_norm": 0.859411358833313, + "learning_rate": 1.0669490874622112e-05, + "loss": 0.1159, + "num_input_tokens_seen": 7414744, + "step": 27655 + }, + { + "epoch": 7.250327653997378, + "grad_norm": 1.8392720222473145, + "learning_rate": 1.0660120651323799e-05, + "loss": 0.1686, + "num_input_tokens_seen": 7415720, + "step": 27660 + }, + { + "epoch": 7.251638269986894, + "grad_norm": 1.4422388076782227, + "learning_rate": 1.0650753429337446e-05, + "loss": 0.241, + "num_input_tokens_seen": 7417160, + "step": 27665 + }, + { + "epoch": 7.252948885976409, + "grad_norm": 1.6585557460784912, + "learning_rate": 1.0641389210623607e-05, + "loss": 0.1937, + "num_input_tokens_seen": 7418696, + "step": 27670 + }, + { + "epoch": 7.254259501965924, + "grad_norm": 0.9300678372383118, + "learning_rate": 1.0632027997142196e-05, + "loss": 0.1605, + "num_input_tokens_seen": 7420264, + "step": 27675 + }, + { + "epoch": 7.255570117955439, + "grad_norm": 2.905836582183838, + "learning_rate": 1.062266979085249e-05, + "loss": 0.2174, + "num_input_tokens_seen": 7421160, + "step": 27680 + }, + { + "epoch": 7.256880733944954, + "grad_norm": 0.8022744655609131, + "learning_rate": 1.0613314593713164e-05, + "loss": 0.1205, + "num_input_tokens_seen": 7422360, + "step": 27685 + }, + { + "epoch": 7.258191349934469, + "grad_norm": 4.8406081199646, + "learning_rate": 1.0603962407682236e-05, + "loss": 0.3324, + "num_input_tokens_seen": 7423720, + "step": 27690 + }, + { + "epoch": 7.259501965923985, + "grad_norm": 1.389313817024231, + "learning_rate": 1.059461323471711e-05, + "loss": 0.1885, + "num_input_tokens_seen": 7425336, + "step": 27695 + }, + { + "epoch": 7.260812581913499, + "grad_norm": 1.5401891469955444, + "learning_rate": 1.0585267076774536e-05, + "loss": 0.3251, + "num_input_tokens_seen": 7426616, + "step": 27700 + }, + { + "epoch": 7.262123197903015, + "grad_norm": 2.6773507595062256, + "learning_rate": 1.0575923935810656e-05, + "loss": 0.1984, + "num_input_tokens_seen": 7428424, + "step": 27705 + }, + { + "epoch": 7.263433813892529, + "grad_norm": 2.863654613494873, + "learning_rate": 1.056658381378097e-05, + "loss": 0.4406, + "num_input_tokens_seen": 7429672, + "step": 27710 + }, + { + "epoch": 7.264744429882045, + "grad_norm": 5.392767429351807, + "learning_rate": 1.0557246712640356e-05, + "loss": 0.1278, + "num_input_tokens_seen": 7430616, + "step": 27715 + }, + { + "epoch": 7.26605504587156, + "grad_norm": 6.985994338989258, + "learning_rate": 1.054791263434305e-05, + "loss": 0.2152, + "num_input_tokens_seen": 7432104, + "step": 27720 + }, + { + "epoch": 7.267365661861074, + "grad_norm": 2.271392583847046, + "learning_rate": 1.0538581580842655e-05, + "loss": 0.1885, + "num_input_tokens_seen": 7433144, + "step": 27725 + }, + { + "epoch": 7.26867627785059, + "grad_norm": 0.5829465985298157, + "learning_rate": 1.052925355409215e-05, + "loss": 0.3291, + "num_input_tokens_seen": 7434664, + "step": 27730 + }, + { + "epoch": 7.269986893840104, + "grad_norm": 0.8773390054702759, + "learning_rate": 1.0519928556043873e-05, + "loss": 0.2386, + "num_input_tokens_seen": 7435624, + "step": 27735 + }, + { + "epoch": 7.27129750982962, + "grad_norm": 2.0049896240234375, + "learning_rate": 1.0510606588649526e-05, + "loss": 0.2357, + "num_input_tokens_seen": 7436840, + "step": 27740 + }, + { + "epoch": 7.272608125819135, + "grad_norm": 1.927771806716919, + "learning_rate": 1.0501287653860189e-05, + "loss": 0.3871, + "num_input_tokens_seen": 7438232, + "step": 27745 + }, + { + "epoch": 7.27391874180865, + "grad_norm": 1.6806918382644653, + "learning_rate": 1.0491971753626292e-05, + "loss": 0.2947, + "num_input_tokens_seen": 7441576, + "step": 27750 + }, + { + "epoch": 7.275229357798165, + "grad_norm": 1.7651770114898682, + "learning_rate": 1.048265888989764e-05, + "loss": 0.2391, + "num_input_tokens_seen": 7443000, + "step": 27755 + }, + { + "epoch": 7.27653997378768, + "grad_norm": 0.904042661190033, + "learning_rate": 1.0473349064623397e-05, + "loss": 0.1385, + "num_input_tokens_seen": 7444488, + "step": 27760 + }, + { + "epoch": 7.277850589777195, + "grad_norm": 1.2356207370758057, + "learning_rate": 1.0464042279752096e-05, + "loss": 0.2647, + "num_input_tokens_seen": 7446088, + "step": 27765 + }, + { + "epoch": 7.2791612057667106, + "grad_norm": 1.863442301750183, + "learning_rate": 1.0454738537231628e-05, + "loss": 0.195, + "num_input_tokens_seen": 7447464, + "step": 27770 + }, + { + "epoch": 7.280471821756225, + "grad_norm": 1.7689927816390991, + "learning_rate": 1.0445437839009265e-05, + "loss": 0.2396, + "num_input_tokens_seen": 7448664, + "step": 27775 + }, + { + "epoch": 7.281782437745741, + "grad_norm": 2.6395275592803955, + "learning_rate": 1.0436140187031601e-05, + "loss": 0.2957, + "num_input_tokens_seen": 7449640, + "step": 27780 + }, + { + "epoch": 7.283093053735255, + "grad_norm": 1.8135836124420166, + "learning_rate": 1.0426845583244632e-05, + "loss": 0.224, + "num_input_tokens_seen": 7450984, + "step": 27785 + }, + { + "epoch": 7.284403669724771, + "grad_norm": 1.9339373111724854, + "learning_rate": 1.04175540295937e-05, + "loss": 0.2083, + "num_input_tokens_seen": 7452200, + "step": 27790 + }, + { + "epoch": 7.285714285714286, + "grad_norm": 2.313613176345825, + "learning_rate": 1.0408265528023508e-05, + "loss": 0.2145, + "num_input_tokens_seen": 7453160, + "step": 27795 + }, + { + "epoch": 7.287024901703801, + "grad_norm": 1.8131109476089478, + "learning_rate": 1.0398980080478139e-05, + "loss": 0.1871, + "num_input_tokens_seen": 7454280, + "step": 27800 + }, + { + "epoch": 7.288335517693316, + "grad_norm": 3.968512535095215, + "learning_rate": 1.0389697688901003e-05, + "loss": 0.272, + "num_input_tokens_seen": 7456184, + "step": 27805 + }, + { + "epoch": 7.289646133682831, + "grad_norm": 3.3252055644989014, + "learning_rate": 1.038041835523491e-05, + "loss": 0.1587, + "num_input_tokens_seen": 7457496, + "step": 27810 + }, + { + "epoch": 7.290956749672346, + "grad_norm": 2.1376841068267822, + "learning_rate": 1.0371142081421987e-05, + "loss": 0.2651, + "num_input_tokens_seen": 7458824, + "step": 27815 + }, + { + "epoch": 7.292267365661861, + "grad_norm": 1.7803252935409546, + "learning_rate": 1.0361868869403745e-05, + "loss": 0.1845, + "num_input_tokens_seen": 7459960, + "step": 27820 + }, + { + "epoch": 7.293577981651376, + "grad_norm": 2.9542253017425537, + "learning_rate": 1.035259872112106e-05, + "loss": 0.3349, + "num_input_tokens_seen": 7461256, + "step": 27825 + }, + { + "epoch": 7.294888597640891, + "grad_norm": 1.8832273483276367, + "learning_rate": 1.0343331638514154e-05, + "loss": 0.1357, + "num_input_tokens_seen": 7462440, + "step": 27830 + }, + { + "epoch": 7.296199213630406, + "grad_norm": 2.7227346897125244, + "learning_rate": 1.0334067623522609e-05, + "loss": 0.0934, + "num_input_tokens_seen": 7463688, + "step": 27835 + }, + { + "epoch": 7.297509829619921, + "grad_norm": 2.3938775062561035, + "learning_rate": 1.032480667808537e-05, + "loss": 0.2306, + "num_input_tokens_seen": 7465096, + "step": 27840 + }, + { + "epoch": 7.2988204456094365, + "grad_norm": 1.0662751197814941, + "learning_rate": 1.0315548804140736e-05, + "loss": 0.1768, + "num_input_tokens_seen": 7466680, + "step": 27845 + }, + { + "epoch": 7.300131061598951, + "grad_norm": 2.023721694946289, + "learning_rate": 1.0306294003626365e-05, + "loss": 0.1358, + "num_input_tokens_seen": 7467768, + "step": 27850 + }, + { + "epoch": 7.301441677588467, + "grad_norm": 2.6800694465637207, + "learning_rate": 1.029704227847927e-05, + "loss": 0.1667, + "num_input_tokens_seen": 7468984, + "step": 27855 + }, + { + "epoch": 7.302752293577981, + "grad_norm": 1.1510251760482788, + "learning_rate": 1.0287793630635822e-05, + "loss": 0.1694, + "num_input_tokens_seen": 7470584, + "step": 27860 + }, + { + "epoch": 7.304062909567497, + "grad_norm": 7.683535575866699, + "learning_rate": 1.0278548062031748e-05, + "loss": 0.2222, + "num_input_tokens_seen": 7472056, + "step": 27865 + }, + { + "epoch": 7.305373525557012, + "grad_norm": 2.7226390838623047, + "learning_rate": 1.0269305574602125e-05, + "loss": 0.172, + "num_input_tokens_seen": 7473576, + "step": 27870 + }, + { + "epoch": 7.306684141546527, + "grad_norm": 1.5707614421844482, + "learning_rate": 1.0260066170281395e-05, + "loss": 0.1201, + "num_input_tokens_seen": 7474856, + "step": 27875 + }, + { + "epoch": 7.307994757536042, + "grad_norm": 0.6885318160057068, + "learning_rate": 1.0250829851003346e-05, + "loss": 0.1915, + "num_input_tokens_seen": 7476392, + "step": 27880 + }, + { + "epoch": 7.309305373525557, + "grad_norm": 1.9577769041061401, + "learning_rate": 1.0241596618701124e-05, + "loss": 0.273, + "num_input_tokens_seen": 7477624, + "step": 27885 + }, + { + "epoch": 7.310615989515072, + "grad_norm": 3.735990047454834, + "learning_rate": 1.0232366475307228e-05, + "loss": 0.2311, + "num_input_tokens_seen": 7478584, + "step": 27890 + }, + { + "epoch": 7.3119266055045875, + "grad_norm": 0.9795308113098145, + "learning_rate": 1.0223139422753524e-05, + "loss": 0.1526, + "num_input_tokens_seen": 7479848, + "step": 27895 + }, + { + "epoch": 7.313237221494102, + "grad_norm": 0.982133686542511, + "learning_rate": 1.0213915462971197e-05, + "loss": 0.2061, + "num_input_tokens_seen": 7481512, + "step": 27900 + }, + { + "epoch": 7.314547837483618, + "grad_norm": 0.24516205489635468, + "learning_rate": 1.0204694597890812e-05, + "loss": 0.2136, + "num_input_tokens_seen": 7482728, + "step": 27905 + }, + { + "epoch": 7.315858453473132, + "grad_norm": 1.7811864614486694, + "learning_rate": 1.0195476829442277e-05, + "loss": 0.2246, + "num_input_tokens_seen": 7483944, + "step": 27910 + }, + { + "epoch": 7.317169069462647, + "grad_norm": 1.574752688407898, + "learning_rate": 1.0186262159554872e-05, + "loss": 0.3531, + "num_input_tokens_seen": 7485608, + "step": 27915 + }, + { + "epoch": 7.3184796854521625, + "grad_norm": 3.5128533840179443, + "learning_rate": 1.0177050590157197e-05, + "loss": 0.167, + "num_input_tokens_seen": 7486728, + "step": 27920 + }, + { + "epoch": 7.319790301441677, + "grad_norm": 2.5499727725982666, + "learning_rate": 1.0167842123177223e-05, + "loss": 0.2035, + "num_input_tokens_seen": 7487912, + "step": 27925 + }, + { + "epoch": 7.321100917431193, + "grad_norm": 1.33086359500885, + "learning_rate": 1.0158636760542276e-05, + "loss": 0.1424, + "num_input_tokens_seen": 7489048, + "step": 27930 + }, + { + "epoch": 7.322411533420707, + "grad_norm": 3.4529001712799072, + "learning_rate": 1.0149434504178997e-05, + "loss": 0.2243, + "num_input_tokens_seen": 7490312, + "step": 27935 + }, + { + "epoch": 7.323722149410223, + "grad_norm": 1.7979310750961304, + "learning_rate": 1.0140235356013422e-05, + "loss": 0.1223, + "num_input_tokens_seen": 7491352, + "step": 27940 + }, + { + "epoch": 7.325032765399738, + "grad_norm": 1.448561668395996, + "learning_rate": 1.0131039317970908e-05, + "loss": 0.1216, + "num_input_tokens_seen": 7492280, + "step": 27945 + }, + { + "epoch": 7.326343381389253, + "grad_norm": 2.1397650241851807, + "learning_rate": 1.0121846391976173e-05, + "loss": 0.3751, + "num_input_tokens_seen": 7493800, + "step": 27950 + }, + { + "epoch": 7.327653997378768, + "grad_norm": 1.0641884803771973, + "learning_rate": 1.0112656579953283e-05, + "loss": 0.1727, + "num_input_tokens_seen": 7495176, + "step": 27955 + }, + { + "epoch": 7.328964613368283, + "grad_norm": 6.542515277862549, + "learning_rate": 1.0103469883825645e-05, + "loss": 0.4218, + "num_input_tokens_seen": 7496472, + "step": 27960 + }, + { + "epoch": 7.330275229357798, + "grad_norm": 2.4811360836029053, + "learning_rate": 1.0094286305516023e-05, + "loss": 0.1761, + "num_input_tokens_seen": 7497752, + "step": 27965 + }, + { + "epoch": 7.3315858453473135, + "grad_norm": 1.1968612670898438, + "learning_rate": 1.008510584694652e-05, + "loss": 0.2032, + "num_input_tokens_seen": 7499000, + "step": 27970 + }, + { + "epoch": 7.332896461336828, + "grad_norm": 1.3075326681137085, + "learning_rate": 1.0075928510038593e-05, + "loss": 0.1141, + "num_input_tokens_seen": 7499992, + "step": 27975 + }, + { + "epoch": 7.334207077326344, + "grad_norm": 0.9253528714179993, + "learning_rate": 1.0066754296713043e-05, + "loss": 0.1367, + "num_input_tokens_seen": 7501192, + "step": 27980 + }, + { + "epoch": 7.335517693315858, + "grad_norm": 1.482942819595337, + "learning_rate": 1.0057583208890017e-05, + "loss": 0.139, + "num_input_tokens_seen": 7502504, + "step": 27985 + }, + { + "epoch": 7.336828309305374, + "grad_norm": 1.405471920967102, + "learning_rate": 1.0048415248489004e-05, + "loss": 0.1818, + "num_input_tokens_seen": 7503944, + "step": 27990 + }, + { + "epoch": 7.3381389252948885, + "grad_norm": 2.602015256881714, + "learning_rate": 1.0039250417428847e-05, + "loss": 0.1791, + "num_input_tokens_seen": 7505240, + "step": 27995 + }, + { + "epoch": 7.339449541284404, + "grad_norm": 1.7551872730255127, + "learning_rate": 1.0030088717627729e-05, + "loss": 0.1993, + "num_input_tokens_seen": 7506344, + "step": 28000 + }, + { + "epoch": 7.340760157273919, + "grad_norm": 1.7383673191070557, + "learning_rate": 1.0020930151003174e-05, + "loss": 0.1155, + "num_input_tokens_seen": 7507688, + "step": 28005 + }, + { + "epoch": 7.342070773263433, + "grad_norm": 5.5293450355529785, + "learning_rate": 1.0011774719472064e-05, + "loss": 0.2441, + "num_input_tokens_seen": 7508760, + "step": 28010 + }, + { + "epoch": 7.343381389252949, + "grad_norm": 1.5019793510437012, + "learning_rate": 1.0002622424950604e-05, + "loss": 0.1519, + "num_input_tokens_seen": 7510136, + "step": 28015 + }, + { + "epoch": 7.344692005242464, + "grad_norm": 0.9664063453674316, + "learning_rate": 9.99347326935435e-06, + "loss": 0.1632, + "num_input_tokens_seen": 7511864, + "step": 28020 + }, + { + "epoch": 7.346002621231979, + "grad_norm": 9.784370422363281, + "learning_rate": 9.984327254598208e-06, + "loss": 0.136, + "num_input_tokens_seen": 7513096, + "step": 28025 + }, + { + "epoch": 7.347313237221494, + "grad_norm": 3.673651695251465, + "learning_rate": 9.975184382596434e-06, + "loss": 0.1587, + "num_input_tokens_seen": 7514136, + "step": 28030 + }, + { + "epoch": 7.348623853211009, + "grad_norm": 1.6477912664413452, + "learning_rate": 9.966044655262607e-06, + "loss": 0.1126, + "num_input_tokens_seen": 7515544, + "step": 28035 + }, + { + "epoch": 7.349934469200524, + "grad_norm": 1.2404471635818481, + "learning_rate": 9.956908074509658e-06, + "loss": 0.1726, + "num_input_tokens_seen": 7516888, + "step": 28040 + }, + { + "epoch": 7.3512450851900395, + "grad_norm": 1.93678617477417, + "learning_rate": 9.947774642249863e-06, + "loss": 0.1383, + "num_input_tokens_seen": 7517944, + "step": 28045 + }, + { + "epoch": 7.352555701179554, + "grad_norm": 2.1731224060058594, + "learning_rate": 9.938644360394822e-06, + "loss": 0.2021, + "num_input_tokens_seen": 7519560, + "step": 28050 + }, + { + "epoch": 7.35386631716907, + "grad_norm": 1.5541207790374756, + "learning_rate": 9.92951723085549e-06, + "loss": 0.1219, + "num_input_tokens_seen": 7520808, + "step": 28055 + }, + { + "epoch": 7.355176933158584, + "grad_norm": 4.454567909240723, + "learning_rate": 9.920393255542163e-06, + "loss": 0.1509, + "num_input_tokens_seen": 7521944, + "step": 28060 + }, + { + "epoch": 7.3564875491481, + "grad_norm": 0.9679951071739197, + "learning_rate": 9.911272436364469e-06, + "loss": 0.208, + "num_input_tokens_seen": 7523464, + "step": 28065 + }, + { + "epoch": 7.3577981651376145, + "grad_norm": 1.6743758916854858, + "learning_rate": 9.902154775231386e-06, + "loss": 0.2006, + "num_input_tokens_seen": 7525304, + "step": 28070 + }, + { + "epoch": 7.35910878112713, + "grad_norm": 1.565550684928894, + "learning_rate": 9.893040274051221e-06, + "loss": 0.1914, + "num_input_tokens_seen": 7526792, + "step": 28075 + }, + { + "epoch": 7.360419397116645, + "grad_norm": 1.3250453472137451, + "learning_rate": 9.883928934731615e-06, + "loss": 0.2899, + "num_input_tokens_seen": 7527864, + "step": 28080 + }, + { + "epoch": 7.36173001310616, + "grad_norm": 4.774748802185059, + "learning_rate": 9.874820759179581e-06, + "loss": 0.2128, + "num_input_tokens_seen": 7529208, + "step": 28085 + }, + { + "epoch": 7.363040629095675, + "grad_norm": 1.7812200784683228, + "learning_rate": 9.865715749301419e-06, + "loss": 0.1866, + "num_input_tokens_seen": 7530696, + "step": 28090 + }, + { + "epoch": 7.3643512450851905, + "grad_norm": 2.2746474742889404, + "learning_rate": 9.8566139070028e-06, + "loss": 0.1459, + "num_input_tokens_seen": 7531800, + "step": 28095 + }, + { + "epoch": 7.365661861074705, + "grad_norm": 1.482304573059082, + "learning_rate": 9.847515234188726e-06, + "loss": 0.2061, + "num_input_tokens_seen": 7532968, + "step": 28100 + }, + { + "epoch": 7.36697247706422, + "grad_norm": 1.7076209783554077, + "learning_rate": 9.838419732763531e-06, + "loss": 0.1245, + "num_input_tokens_seen": 7534264, + "step": 28105 + }, + { + "epoch": 7.368283093053735, + "grad_norm": 1.489607334136963, + "learning_rate": 9.829327404630891e-06, + "loss": 0.1856, + "num_input_tokens_seen": 7535720, + "step": 28110 + }, + { + "epoch": 7.36959370904325, + "grad_norm": 1.2970558404922485, + "learning_rate": 9.820238251693811e-06, + "loss": 0.255, + "num_input_tokens_seen": 7536792, + "step": 28115 + }, + { + "epoch": 7.3709043250327655, + "grad_norm": 0.8919407725334167, + "learning_rate": 9.811152275854635e-06, + "loss": 0.165, + "num_input_tokens_seen": 7537864, + "step": 28120 + }, + { + "epoch": 7.37221494102228, + "grad_norm": 1.0322389602661133, + "learning_rate": 9.802069479015044e-06, + "loss": 0.2255, + "num_input_tokens_seen": 7539272, + "step": 28125 + }, + { + "epoch": 7.373525557011796, + "grad_norm": 6.315885066986084, + "learning_rate": 9.792989863076064e-06, + "loss": 0.2841, + "num_input_tokens_seen": 7540296, + "step": 28130 + }, + { + "epoch": 7.37483617300131, + "grad_norm": 2.022188901901245, + "learning_rate": 9.783913429938015e-06, + "loss": 0.2792, + "num_input_tokens_seen": 7541768, + "step": 28135 + }, + { + "epoch": 7.376146788990826, + "grad_norm": 2.059007406234741, + "learning_rate": 9.774840181500588e-06, + "loss": 0.1949, + "num_input_tokens_seen": 7542952, + "step": 28140 + }, + { + "epoch": 7.3774574049803405, + "grad_norm": 2.6216018199920654, + "learning_rate": 9.765770119662813e-06, + "loss": 0.1595, + "num_input_tokens_seen": 7543912, + "step": 28145 + }, + { + "epoch": 7.378768020969856, + "grad_norm": 1.653842568397522, + "learning_rate": 9.756703246323026e-06, + "loss": 0.2023, + "num_input_tokens_seen": 7544856, + "step": 28150 + }, + { + "epoch": 7.380078636959371, + "grad_norm": 3.69905161857605, + "learning_rate": 9.747639563378916e-06, + "loss": 0.1786, + "num_input_tokens_seen": 7545992, + "step": 28155 + }, + { + "epoch": 7.381389252948886, + "grad_norm": 1.3095831871032715, + "learning_rate": 9.738579072727488e-06, + "loss": 0.1745, + "num_input_tokens_seen": 7547496, + "step": 28160 + }, + { + "epoch": 7.382699868938401, + "grad_norm": 1.7067135572433472, + "learning_rate": 9.7295217762651e-06, + "loss": 0.2385, + "num_input_tokens_seen": 7548552, + "step": 28165 + }, + { + "epoch": 7.3840104849279165, + "grad_norm": 0.8404866456985474, + "learning_rate": 9.72046767588741e-06, + "loss": 0.2788, + "num_input_tokens_seen": 7550280, + "step": 28170 + }, + { + "epoch": 7.385321100917431, + "grad_norm": 2.409482479095459, + "learning_rate": 9.711416773489434e-06, + "loss": 0.1578, + "num_input_tokens_seen": 7551368, + "step": 28175 + }, + { + "epoch": 7.386631716906947, + "grad_norm": 8.890357971191406, + "learning_rate": 9.702369070965512e-06, + "loss": 0.2236, + "num_input_tokens_seen": 7552808, + "step": 28180 + }, + { + "epoch": 7.387942332896461, + "grad_norm": 3.5052871704101562, + "learning_rate": 9.693324570209312e-06, + "loss": 0.1668, + "num_input_tokens_seen": 7554072, + "step": 28185 + }, + { + "epoch": 7.389252948885977, + "grad_norm": 3.103571891784668, + "learning_rate": 9.684283273113829e-06, + "loss": 0.3536, + "num_input_tokens_seen": 7555336, + "step": 28190 + }, + { + "epoch": 7.3905635648754915, + "grad_norm": 1.5666157007217407, + "learning_rate": 9.675245181571385e-06, + "loss": 0.0981, + "num_input_tokens_seen": 7556536, + "step": 28195 + }, + { + "epoch": 7.391874180865006, + "grad_norm": 2.9337496757507324, + "learning_rate": 9.666210297473663e-06, + "loss": 0.2001, + "num_input_tokens_seen": 7557592, + "step": 28200 + }, + { + "epoch": 7.393184796854522, + "grad_norm": 0.6337182521820068, + "learning_rate": 9.657178622711624e-06, + "loss": 0.1113, + "num_input_tokens_seen": 7558808, + "step": 28205 + }, + { + "epoch": 7.394495412844036, + "grad_norm": 1.6782288551330566, + "learning_rate": 9.648150159175587e-06, + "loss": 0.1483, + "num_input_tokens_seen": 7559992, + "step": 28210 + }, + { + "epoch": 7.395806028833552, + "grad_norm": 1.0374641418457031, + "learning_rate": 9.639124908755196e-06, + "loss": 0.1676, + "num_input_tokens_seen": 7561560, + "step": 28215 + }, + { + "epoch": 7.3971166448230665, + "grad_norm": 1.2471014261245728, + "learning_rate": 9.630102873339419e-06, + "loss": 0.1342, + "num_input_tokens_seen": 7563000, + "step": 28220 + }, + { + "epoch": 7.398427260812582, + "grad_norm": 1.89713454246521, + "learning_rate": 9.621084054816557e-06, + "loss": 0.1716, + "num_input_tokens_seen": 7564664, + "step": 28225 + }, + { + "epoch": 7.399737876802097, + "grad_norm": 1.0666497945785522, + "learning_rate": 9.612068455074227e-06, + "loss": 0.1514, + "num_input_tokens_seen": 7566056, + "step": 28230 + }, + { + "epoch": 7.401048492791612, + "grad_norm": 1.7268072366714478, + "learning_rate": 9.603056075999382e-06, + "loss": 0.3266, + "num_input_tokens_seen": 7567288, + "step": 28235 + }, + { + "epoch": 7.402359108781127, + "grad_norm": 5.543944358825684, + "learning_rate": 9.594046919478294e-06, + "loss": 0.3235, + "num_input_tokens_seen": 7568824, + "step": 28240 + }, + { + "epoch": 7.4036697247706424, + "grad_norm": 2.417271614074707, + "learning_rate": 9.58504098739658e-06, + "loss": 0.2551, + "num_input_tokens_seen": 7570504, + "step": 28245 + }, + { + "epoch": 7.404980340760157, + "grad_norm": 1.3268566131591797, + "learning_rate": 9.576038281639138e-06, + "loss": 0.2206, + "num_input_tokens_seen": 7571784, + "step": 28250 + }, + { + "epoch": 7.406290956749673, + "grad_norm": 2.5505330562591553, + "learning_rate": 9.567038804090227e-06, + "loss": 0.2567, + "num_input_tokens_seen": 7573112, + "step": 28255 + }, + { + "epoch": 7.407601572739187, + "grad_norm": 1.8189477920532227, + "learning_rate": 9.55804255663344e-06, + "loss": 0.242, + "num_input_tokens_seen": 7574536, + "step": 28260 + }, + { + "epoch": 7.408912188728703, + "grad_norm": 1.7084472179412842, + "learning_rate": 9.549049541151659e-06, + "loss": 0.2277, + "num_input_tokens_seen": 7575848, + "step": 28265 + }, + { + "epoch": 7.4102228047182175, + "grad_norm": 14.696775436401367, + "learning_rate": 9.540059759527115e-06, + "loss": 0.2386, + "num_input_tokens_seen": 7577016, + "step": 28270 + }, + { + "epoch": 7.411533420707733, + "grad_norm": 1.426666021347046, + "learning_rate": 9.53107321364135e-06, + "loss": 0.1807, + "num_input_tokens_seen": 7578488, + "step": 28275 + }, + { + "epoch": 7.412844036697248, + "grad_norm": 1.8475298881530762, + "learning_rate": 9.522089905375231e-06, + "loss": 0.2544, + "num_input_tokens_seen": 7579960, + "step": 28280 + }, + { + "epoch": 7.414154652686763, + "grad_norm": 1.4163480997085571, + "learning_rate": 9.513109836608963e-06, + "loss": 0.1058, + "num_input_tokens_seen": 7581432, + "step": 28285 + }, + { + "epoch": 7.415465268676278, + "grad_norm": 0.7123882174491882, + "learning_rate": 9.504133009222035e-06, + "loss": 0.1817, + "num_input_tokens_seen": 7582552, + "step": 28290 + }, + { + "epoch": 7.4167758846657925, + "grad_norm": 1.2037783861160278, + "learning_rate": 9.495159425093298e-06, + "loss": 0.1553, + "num_input_tokens_seen": 7583944, + "step": 28295 + }, + { + "epoch": 7.418086500655308, + "grad_norm": 2.6128342151641846, + "learning_rate": 9.4861890861009e-06, + "loss": 0.2329, + "num_input_tokens_seen": 7585320, + "step": 28300 + }, + { + "epoch": 7.419397116644823, + "grad_norm": 21.960044860839844, + "learning_rate": 9.477221994122321e-06, + "loss": 0.2881, + "num_input_tokens_seen": 7586456, + "step": 28305 + }, + { + "epoch": 7.420707732634338, + "grad_norm": 0.4782097339630127, + "learning_rate": 9.468258151034349e-06, + "loss": 0.2024, + "num_input_tokens_seen": 7587560, + "step": 28310 + }, + { + "epoch": 7.422018348623853, + "grad_norm": 1.8849984407424927, + "learning_rate": 9.459297558713115e-06, + "loss": 0.2976, + "num_input_tokens_seen": 7588664, + "step": 28315 + }, + { + "epoch": 7.423328964613368, + "grad_norm": 1.4209730625152588, + "learning_rate": 9.45034021903406e-06, + "loss": 0.2961, + "num_input_tokens_seen": 7589848, + "step": 28320 + }, + { + "epoch": 7.424639580602883, + "grad_norm": 1.8454806804656982, + "learning_rate": 9.441386133871922e-06, + "loss": 0.319, + "num_input_tokens_seen": 7590920, + "step": 28325 + }, + { + "epoch": 7.425950196592399, + "grad_norm": 3.1632938385009766, + "learning_rate": 9.432435305100782e-06, + "loss": 0.2633, + "num_input_tokens_seen": 7592232, + "step": 28330 + }, + { + "epoch": 7.427260812581913, + "grad_norm": 1.3682037591934204, + "learning_rate": 9.423487734594036e-06, + "loss": 0.1862, + "num_input_tokens_seen": 7593624, + "step": 28335 + }, + { + "epoch": 7.428571428571429, + "grad_norm": 1.2510695457458496, + "learning_rate": 9.41454342422439e-06, + "loss": 0.2588, + "num_input_tokens_seen": 7595096, + "step": 28340 + }, + { + "epoch": 7.4298820445609435, + "grad_norm": 2.554290771484375, + "learning_rate": 9.405602375863876e-06, + "loss": 0.1975, + "num_input_tokens_seen": 7596664, + "step": 28345 + }, + { + "epoch": 7.431192660550459, + "grad_norm": 2.330089807510376, + "learning_rate": 9.39666459138384e-06, + "loss": 0.1713, + "num_input_tokens_seen": 7597624, + "step": 28350 + }, + { + "epoch": 7.432503276539974, + "grad_norm": 2.052928924560547, + "learning_rate": 9.387730072654946e-06, + "loss": 0.2336, + "num_input_tokens_seen": 7598984, + "step": 28355 + }, + { + "epoch": 7.433813892529489, + "grad_norm": 2.243804931640625, + "learning_rate": 9.378798821547171e-06, + "loss": 0.3892, + "num_input_tokens_seen": 7600216, + "step": 28360 + }, + { + "epoch": 7.435124508519004, + "grad_norm": 2.5944509506225586, + "learning_rate": 9.36987083992982e-06, + "loss": 0.2003, + "num_input_tokens_seen": 7601496, + "step": 28365 + }, + { + "epoch": 7.436435124508519, + "grad_norm": 1.8712255954742432, + "learning_rate": 9.36094612967148e-06, + "loss": 0.3527, + "num_input_tokens_seen": 7603400, + "step": 28370 + }, + { + "epoch": 7.437745740498034, + "grad_norm": 2.514359712600708, + "learning_rate": 9.352024692640098e-06, + "loss": 0.1718, + "num_input_tokens_seen": 7604632, + "step": 28375 + }, + { + "epoch": 7.43905635648755, + "grad_norm": 2.2774288654327393, + "learning_rate": 9.343106530702913e-06, + "loss": 0.1599, + "num_input_tokens_seen": 7605672, + "step": 28380 + }, + { + "epoch": 7.440366972477064, + "grad_norm": 1.8112581968307495, + "learning_rate": 9.334191645726475e-06, + "loss": 0.166, + "num_input_tokens_seen": 7606856, + "step": 28385 + }, + { + "epoch": 7.441677588466579, + "grad_norm": 1.5417684316635132, + "learning_rate": 9.325280039576661e-06, + "loss": 0.219, + "num_input_tokens_seen": 7608120, + "step": 28390 + }, + { + "epoch": 7.442988204456094, + "grad_norm": 1.6360572576522827, + "learning_rate": 9.31637171411865e-06, + "loss": 0.1425, + "num_input_tokens_seen": 7609224, + "step": 28395 + }, + { + "epoch": 7.444298820445609, + "grad_norm": 1.5032660961151123, + "learning_rate": 9.307466671216946e-06, + "loss": 0.2925, + "num_input_tokens_seen": 7610520, + "step": 28400 + }, + { + "epoch": 7.445609436435125, + "grad_norm": 4.702399253845215, + "learning_rate": 9.298564912735347e-06, + "loss": 0.2493, + "num_input_tokens_seen": 7612328, + "step": 28405 + }, + { + "epoch": 7.446920052424639, + "grad_norm": 1.2686188220977783, + "learning_rate": 9.289666440536981e-06, + "loss": 0.2172, + "num_input_tokens_seen": 7613784, + "step": 28410 + }, + { + "epoch": 7.448230668414155, + "grad_norm": 1.8338749408721924, + "learning_rate": 9.280771256484284e-06, + "loss": 0.3083, + "num_input_tokens_seen": 7615224, + "step": 28415 + }, + { + "epoch": 7.4495412844036695, + "grad_norm": 3.2806780338287354, + "learning_rate": 9.271879362439e-06, + "loss": 0.191, + "num_input_tokens_seen": 7616600, + "step": 28420 + }, + { + "epoch": 7.450851900393185, + "grad_norm": 3.3530240058898926, + "learning_rate": 9.262990760262183e-06, + "loss": 0.1163, + "num_input_tokens_seen": 7617720, + "step": 28425 + }, + { + "epoch": 7.4521625163827, + "grad_norm": 1.6026439666748047, + "learning_rate": 9.254105451814215e-06, + "loss": 0.2161, + "num_input_tokens_seen": 7619432, + "step": 28430 + }, + { + "epoch": 7.453473132372215, + "grad_norm": 2.468729019165039, + "learning_rate": 9.245223438954781e-06, + "loss": 0.1713, + "num_input_tokens_seen": 7620408, + "step": 28435 + }, + { + "epoch": 7.45478374836173, + "grad_norm": 1.3474754095077515, + "learning_rate": 9.236344723542847e-06, + "loss": 0.1316, + "num_input_tokens_seen": 7621624, + "step": 28440 + }, + { + "epoch": 7.456094364351245, + "grad_norm": 2.400163173675537, + "learning_rate": 9.227469307436728e-06, + "loss": 0.152, + "num_input_tokens_seen": 7623080, + "step": 28445 + }, + { + "epoch": 7.45740498034076, + "grad_norm": 3.359159231185913, + "learning_rate": 9.21859719249403e-06, + "loss": 0.1525, + "num_input_tokens_seen": 7624120, + "step": 28450 + }, + { + "epoch": 7.458715596330276, + "grad_norm": 1.0885778665542603, + "learning_rate": 9.209728380571669e-06, + "loss": 0.1754, + "num_input_tokens_seen": 7625496, + "step": 28455 + }, + { + "epoch": 7.46002621231979, + "grad_norm": 0.769768238067627, + "learning_rate": 9.200862873525876e-06, + "loss": 0.2337, + "num_input_tokens_seen": 7627656, + "step": 28460 + }, + { + "epoch": 7.461336828309306, + "grad_norm": 1.6548768281936646, + "learning_rate": 9.192000673212184e-06, + "loss": 0.1826, + "num_input_tokens_seen": 7629000, + "step": 28465 + }, + { + "epoch": 7.46264744429882, + "grad_norm": 2.626798391342163, + "learning_rate": 9.183141781485439e-06, + "loss": 0.1923, + "num_input_tokens_seen": 7630200, + "step": 28470 + }, + { + "epoch": 7.463958060288336, + "grad_norm": 2.3668811321258545, + "learning_rate": 9.174286200199786e-06, + "loss": 0.209, + "num_input_tokens_seen": 7631208, + "step": 28475 + }, + { + "epoch": 7.465268676277851, + "grad_norm": 2.8484015464782715, + "learning_rate": 9.165433931208697e-06, + "loss": 0.222, + "num_input_tokens_seen": 7632360, + "step": 28480 + }, + { + "epoch": 7.466579292267365, + "grad_norm": 1.776762843132019, + "learning_rate": 9.15658497636491e-06, + "loss": 0.2613, + "num_input_tokens_seen": 7633704, + "step": 28485 + }, + { + "epoch": 7.467889908256881, + "grad_norm": 2.035057306289673, + "learning_rate": 9.147739337520519e-06, + "loss": 0.1635, + "num_input_tokens_seen": 7634776, + "step": 28490 + }, + { + "epoch": 7.4692005242463955, + "grad_norm": 9.28377628326416, + "learning_rate": 9.138897016526895e-06, + "loss": 0.3299, + "num_input_tokens_seen": 7636456, + "step": 28495 + }, + { + "epoch": 7.470511140235911, + "grad_norm": 1.732108235359192, + "learning_rate": 9.130058015234722e-06, + "loss": 0.2708, + "num_input_tokens_seen": 7637704, + "step": 28500 + }, + { + "epoch": 7.471821756225426, + "grad_norm": 1.4381345510482788, + "learning_rate": 9.12122233549398e-06, + "loss": 0.1146, + "num_input_tokens_seen": 7639000, + "step": 28505 + }, + { + "epoch": 7.473132372214941, + "grad_norm": 1.9797694683074951, + "learning_rate": 9.11238997915397e-06, + "loss": 0.1682, + "num_input_tokens_seen": 7640200, + "step": 28510 + }, + { + "epoch": 7.474442988204456, + "grad_norm": 2.6152467727661133, + "learning_rate": 9.103560948063286e-06, + "loss": 0.2563, + "num_input_tokens_seen": 7641640, + "step": 28515 + }, + { + "epoch": 7.475753604193971, + "grad_norm": 1.9442487955093384, + "learning_rate": 9.094735244069839e-06, + "loss": 0.1207, + "num_input_tokens_seen": 7642872, + "step": 28520 + }, + { + "epoch": 7.477064220183486, + "grad_norm": 1.3032090663909912, + "learning_rate": 9.08591286902081e-06, + "loss": 0.2718, + "num_input_tokens_seen": 7644456, + "step": 28525 + }, + { + "epoch": 7.478374836173002, + "grad_norm": 2.260958433151245, + "learning_rate": 9.077093824762722e-06, + "loss": 0.207, + "num_input_tokens_seen": 7645592, + "step": 28530 + }, + { + "epoch": 7.479685452162516, + "grad_norm": 1.3643990755081177, + "learning_rate": 9.068278113141385e-06, + "loss": 0.1083, + "num_input_tokens_seen": 7646744, + "step": 28535 + }, + { + "epoch": 7.480996068152032, + "grad_norm": 1.6546279191970825, + "learning_rate": 9.059465736001901e-06, + "loss": 0.2862, + "num_input_tokens_seen": 7648392, + "step": 28540 + }, + { + "epoch": 7.482306684141546, + "grad_norm": 1.9233704805374146, + "learning_rate": 9.050656695188706e-06, + "loss": 0.244, + "num_input_tokens_seen": 7650344, + "step": 28545 + }, + { + "epoch": 7.483617300131062, + "grad_norm": 1.906860113143921, + "learning_rate": 9.041850992545503e-06, + "loss": 0.1656, + "num_input_tokens_seen": 7651544, + "step": 28550 + }, + { + "epoch": 7.484927916120577, + "grad_norm": 2.6731157302856445, + "learning_rate": 9.033048629915326e-06, + "loss": 0.1908, + "num_input_tokens_seen": 7653160, + "step": 28555 + }, + { + "epoch": 7.486238532110092, + "grad_norm": 1.8495068550109863, + "learning_rate": 9.02424960914047e-06, + "loss": 0.1111, + "num_input_tokens_seen": 7654152, + "step": 28560 + }, + { + "epoch": 7.487549148099607, + "grad_norm": 1.852125883102417, + "learning_rate": 9.015453932062568e-06, + "loss": 0.1187, + "num_input_tokens_seen": 7655560, + "step": 28565 + }, + { + "epoch": 7.488859764089122, + "grad_norm": 3.690875291824341, + "learning_rate": 9.006661600522539e-06, + "loss": 0.3093, + "num_input_tokens_seen": 7656744, + "step": 28570 + }, + { + "epoch": 7.490170380078637, + "grad_norm": 3.3652331829071045, + "learning_rate": 8.997872616360603e-06, + "loss": 0.3119, + "num_input_tokens_seen": 7658216, + "step": 28575 + }, + { + "epoch": 7.491480996068152, + "grad_norm": 1.196881651878357, + "learning_rate": 8.989086981416278e-06, + "loss": 0.2344, + "num_input_tokens_seen": 7660504, + "step": 28580 + }, + { + "epoch": 7.492791612057667, + "grad_norm": 2.221950054168701, + "learning_rate": 8.980304697528385e-06, + "loss": 0.2531, + "num_input_tokens_seen": 7661720, + "step": 28585 + }, + { + "epoch": 7.494102228047182, + "grad_norm": 2.020498275756836, + "learning_rate": 8.971525766535038e-06, + "loss": 0.2914, + "num_input_tokens_seen": 7663448, + "step": 28590 + }, + { + "epoch": 7.495412844036697, + "grad_norm": 2.740054130554199, + "learning_rate": 8.962750190273653e-06, + "loss": 0.3201, + "num_input_tokens_seen": 7665048, + "step": 28595 + }, + { + "epoch": 7.496723460026212, + "grad_norm": 1.69959557056427, + "learning_rate": 8.953977970580942e-06, + "loss": 0.1528, + "num_input_tokens_seen": 7666600, + "step": 28600 + }, + { + "epoch": 7.498034076015728, + "grad_norm": 1.1631224155426025, + "learning_rate": 8.945209109292919e-06, + "loss": 0.2277, + "num_input_tokens_seen": 7667960, + "step": 28605 + }, + { + "epoch": 7.499344692005242, + "grad_norm": 0.9878867268562317, + "learning_rate": 8.936443608244885e-06, + "loss": 0.195, + "num_input_tokens_seen": 7669336, + "step": 28610 + }, + { + "epoch": 7.500655307994758, + "grad_norm": 3.7121942043304443, + "learning_rate": 8.927681469271451e-06, + "loss": 0.2152, + "num_input_tokens_seen": 7670600, + "step": 28615 + }, + { + "epoch": 7.501965923984272, + "grad_norm": 1.5253578424453735, + "learning_rate": 8.918922694206513e-06, + "loss": 0.2348, + "num_input_tokens_seen": 7672088, + "step": 28620 + }, + { + "epoch": 7.501965923984272, + "eval_loss": 0.7353919148445129, + "eval_runtime": 16.8748, + "eval_samples_per_second": 50.252, + "eval_steps_per_second": 25.126, + "num_input_tokens_seen": 7672088, + "step": 28620 + }, + { + "epoch": 7.503276539973788, + "grad_norm": 1.6474976539611816, + "learning_rate": 8.91016728488327e-06, + "loss": 0.2059, + "num_input_tokens_seen": 7673528, + "step": 28625 + }, + { + "epoch": 7.504587155963303, + "grad_norm": 1.5424978733062744, + "learning_rate": 8.901415243134211e-06, + "loss": 0.1385, + "num_input_tokens_seen": 7674808, + "step": 28630 + }, + { + "epoch": 7.505897771952818, + "grad_norm": 1.830735683441162, + "learning_rate": 8.892666570791136e-06, + "loss": 0.1382, + "num_input_tokens_seen": 7676008, + "step": 28635 + }, + { + "epoch": 7.507208387942333, + "grad_norm": 3.114382028579712, + "learning_rate": 8.883921269685108e-06, + "loss": 0.2967, + "num_input_tokens_seen": 7677736, + "step": 28640 + }, + { + "epoch": 7.508519003931848, + "grad_norm": 1.9076248407363892, + "learning_rate": 8.875179341646511e-06, + "loss": 0.2796, + "num_input_tokens_seen": 7679192, + "step": 28645 + }, + { + "epoch": 7.509829619921363, + "grad_norm": 2.214456796646118, + "learning_rate": 8.866440788505018e-06, + "loss": 0.1383, + "num_input_tokens_seen": 7680600, + "step": 28650 + }, + { + "epoch": 7.511140235910878, + "grad_norm": 1.3632149696350098, + "learning_rate": 8.857705612089581e-06, + "loss": 0.2365, + "num_input_tokens_seen": 7681992, + "step": 28655 + }, + { + "epoch": 7.512450851900393, + "grad_norm": 2.3220036029815674, + "learning_rate": 8.848973814228478e-06, + "loss": 0.1568, + "num_input_tokens_seen": 7683112, + "step": 28660 + }, + { + "epoch": 7.513761467889909, + "grad_norm": 1.0480046272277832, + "learning_rate": 8.840245396749247e-06, + "loss": 0.182, + "num_input_tokens_seen": 7684440, + "step": 28665 + }, + { + "epoch": 7.515072083879423, + "grad_norm": 1.2526289224624634, + "learning_rate": 8.831520361478744e-06, + "loss": 0.2084, + "num_input_tokens_seen": 7686264, + "step": 28670 + }, + { + "epoch": 7.516382699868938, + "grad_norm": 1.3697770833969116, + "learning_rate": 8.82279871024308e-06, + "loss": 0.1396, + "num_input_tokens_seen": 7687688, + "step": 28675 + }, + { + "epoch": 7.517693315858454, + "grad_norm": 2.8271849155426025, + "learning_rate": 8.814080444867695e-06, + "loss": 0.3554, + "num_input_tokens_seen": 7690104, + "step": 28680 + }, + { + "epoch": 7.519003931847968, + "grad_norm": 0.9918498396873474, + "learning_rate": 8.805365567177307e-06, + "loss": 0.1457, + "num_input_tokens_seen": 7691544, + "step": 28685 + }, + { + "epoch": 7.520314547837484, + "grad_norm": 2.276144504547119, + "learning_rate": 8.796654078995921e-06, + "loss": 0.1052, + "num_input_tokens_seen": 7692712, + "step": 28690 + }, + { + "epoch": 7.521625163826998, + "grad_norm": 1.561903476715088, + "learning_rate": 8.787945982146841e-06, + "loss": 0.1591, + "num_input_tokens_seen": 7693736, + "step": 28695 + }, + { + "epoch": 7.522935779816514, + "grad_norm": 0.5061004161834717, + "learning_rate": 8.779241278452657e-06, + "loss": 0.1529, + "num_input_tokens_seen": 7695032, + "step": 28700 + }, + { + "epoch": 7.524246395806029, + "grad_norm": 2.0124287605285645, + "learning_rate": 8.770539969735241e-06, + "loss": 0.1548, + "num_input_tokens_seen": 7696216, + "step": 28705 + }, + { + "epoch": 7.525557011795544, + "grad_norm": 2.4093568325042725, + "learning_rate": 8.761842057815772e-06, + "loss": 0.2747, + "num_input_tokens_seen": 7697720, + "step": 28710 + }, + { + "epoch": 7.526867627785059, + "grad_norm": 2.6331679821014404, + "learning_rate": 8.753147544514701e-06, + "loss": 0.1229, + "num_input_tokens_seen": 7699144, + "step": 28715 + }, + { + "epoch": 7.528178243774574, + "grad_norm": 1.2461206912994385, + "learning_rate": 8.744456431651777e-06, + "loss": 0.1801, + "num_input_tokens_seen": 7700440, + "step": 28720 + }, + { + "epoch": 7.529488859764089, + "grad_norm": 2.0356345176696777, + "learning_rate": 8.73576872104603e-06, + "loss": 0.1806, + "num_input_tokens_seen": 7701624, + "step": 28725 + }, + { + "epoch": 7.5307994757536045, + "grad_norm": 3.5184426307678223, + "learning_rate": 8.727084414515793e-06, + "loss": 0.1854, + "num_input_tokens_seen": 7702600, + "step": 28730 + }, + { + "epoch": 7.532110091743119, + "grad_norm": 0.20871560275554657, + "learning_rate": 8.718403513878668e-06, + "loss": 0.1056, + "num_input_tokens_seen": 7703672, + "step": 28735 + }, + { + "epoch": 7.533420707732635, + "grad_norm": 2.4455807209014893, + "learning_rate": 8.709726020951557e-06, + "loss": 0.1901, + "num_input_tokens_seen": 7704776, + "step": 28740 + }, + { + "epoch": 7.534731323722149, + "grad_norm": 2.737389326095581, + "learning_rate": 8.701051937550638e-06, + "loss": 0.2819, + "num_input_tokens_seen": 7706328, + "step": 28745 + }, + { + "epoch": 7.536041939711664, + "grad_norm": 2.9182655811309814, + "learning_rate": 8.69238126549139e-06, + "loss": 0.1218, + "num_input_tokens_seen": 7707416, + "step": 28750 + }, + { + "epoch": 7.53735255570118, + "grad_norm": 3.625020742416382, + "learning_rate": 8.683714006588576e-06, + "loss": 0.2446, + "num_input_tokens_seen": 7708696, + "step": 28755 + }, + { + "epoch": 7.538663171690695, + "grad_norm": 0.3332306742668152, + "learning_rate": 8.675050162656215e-06, + "loss": 0.1938, + "num_input_tokens_seen": 7709992, + "step": 28760 + }, + { + "epoch": 7.53997378768021, + "grad_norm": 2.1550676822662354, + "learning_rate": 8.666389735507648e-06, + "loss": 0.1649, + "num_input_tokens_seen": 7710888, + "step": 28765 + }, + { + "epoch": 7.541284403669724, + "grad_norm": 1.5309818983078003, + "learning_rate": 8.65773272695548e-06, + "loss": 0.1843, + "num_input_tokens_seen": 7711960, + "step": 28770 + }, + { + "epoch": 7.54259501965924, + "grad_norm": 1.3728115558624268, + "learning_rate": 8.649079138811625e-06, + "loss": 0.2414, + "num_input_tokens_seen": 7713160, + "step": 28775 + }, + { + "epoch": 7.543905635648755, + "grad_norm": 1.4187183380126953, + "learning_rate": 8.640428972887252e-06, + "loss": 0.2081, + "num_input_tokens_seen": 7714792, + "step": 28780 + }, + { + "epoch": 7.54521625163827, + "grad_norm": 16.03165054321289, + "learning_rate": 8.63178223099283e-06, + "loss": 0.1986, + "num_input_tokens_seen": 7715896, + "step": 28785 + }, + { + "epoch": 7.546526867627785, + "grad_norm": 2.423797130584717, + "learning_rate": 8.623138914938111e-06, + "loss": 0.2176, + "num_input_tokens_seen": 7717144, + "step": 28790 + }, + { + "epoch": 7.5478374836173, + "grad_norm": 1.3603341579437256, + "learning_rate": 8.61449902653211e-06, + "loss": 0.2206, + "num_input_tokens_seen": 7718488, + "step": 28795 + }, + { + "epoch": 7.549148099606815, + "grad_norm": 2.1876864433288574, + "learning_rate": 8.605862567583155e-06, + "loss": 0.1621, + "num_input_tokens_seen": 7720312, + "step": 28800 + }, + { + "epoch": 7.5504587155963305, + "grad_norm": 3.9942233562469482, + "learning_rate": 8.597229539898835e-06, + "loss": 0.2255, + "num_input_tokens_seen": 7721432, + "step": 28805 + }, + { + "epoch": 7.551769331585845, + "grad_norm": 1.8647918701171875, + "learning_rate": 8.588599945286033e-06, + "loss": 0.1365, + "num_input_tokens_seen": 7722968, + "step": 28810 + }, + { + "epoch": 7.553079947575361, + "grad_norm": 2.1478025913238525, + "learning_rate": 8.579973785550904e-06, + "loss": 0.211, + "num_input_tokens_seen": 7724280, + "step": 28815 + }, + { + "epoch": 7.554390563564875, + "grad_norm": 2.091050624847412, + "learning_rate": 8.571351062498888e-06, + "loss": 0.1485, + "num_input_tokens_seen": 7725256, + "step": 28820 + }, + { + "epoch": 7.555701179554391, + "grad_norm": 3.187635898590088, + "learning_rate": 8.562731777934713e-06, + "loss": 0.1818, + "num_input_tokens_seen": 7726168, + "step": 28825 + }, + { + "epoch": 7.557011795543906, + "grad_norm": 2.876556873321533, + "learning_rate": 8.55411593366237e-06, + "loss": 0.1572, + "num_input_tokens_seen": 7727416, + "step": 28830 + }, + { + "epoch": 7.558322411533421, + "grad_norm": 1.2040342092514038, + "learning_rate": 8.545503531485149e-06, + "loss": 0.0849, + "num_input_tokens_seen": 7728600, + "step": 28835 + }, + { + "epoch": 7.559633027522936, + "grad_norm": 3.1685774326324463, + "learning_rate": 8.536894573205608e-06, + "loss": 0.3689, + "num_input_tokens_seen": 7730008, + "step": 28840 + }, + { + "epoch": 7.56094364351245, + "grad_norm": 2.660550832748413, + "learning_rate": 8.528289060625586e-06, + "loss": 0.2694, + "num_input_tokens_seen": 7731112, + "step": 28845 + }, + { + "epoch": 7.562254259501966, + "grad_norm": 2.1907169818878174, + "learning_rate": 8.519686995546203e-06, + "loss": 0.2072, + "num_input_tokens_seen": 7732376, + "step": 28850 + }, + { + "epoch": 7.5635648754914815, + "grad_norm": 3.70902419090271, + "learning_rate": 8.511088379767857e-06, + "loss": 0.1137, + "num_input_tokens_seen": 7733640, + "step": 28855 + }, + { + "epoch": 7.564875491480996, + "grad_norm": 1.2953143119812012, + "learning_rate": 8.502493215090218e-06, + "loss": 0.2554, + "num_input_tokens_seen": 7735064, + "step": 28860 + }, + { + "epoch": 7.566186107470511, + "grad_norm": 1.7317169904708862, + "learning_rate": 8.493901503312249e-06, + "loss": 0.2417, + "num_input_tokens_seen": 7736328, + "step": 28865 + }, + { + "epoch": 7.567496723460026, + "grad_norm": 1.6235301494598389, + "learning_rate": 8.48531324623218e-06, + "loss": 0.1219, + "num_input_tokens_seen": 7737672, + "step": 28870 + }, + { + "epoch": 7.568807339449541, + "grad_norm": 1.5818941593170166, + "learning_rate": 8.476728445647506e-06, + "loss": 0.178, + "num_input_tokens_seen": 7739272, + "step": 28875 + }, + { + "epoch": 7.5701179554390565, + "grad_norm": 3.46589732170105, + "learning_rate": 8.468147103355018e-06, + "loss": 0.2379, + "num_input_tokens_seen": 7740376, + "step": 28880 + }, + { + "epoch": 7.571428571428571, + "grad_norm": 3.973001718521118, + "learning_rate": 8.45956922115077e-06, + "loss": 0.2454, + "num_input_tokens_seen": 7741816, + "step": 28885 + }, + { + "epoch": 7.572739187418087, + "grad_norm": 1.1320160627365112, + "learning_rate": 8.450994800830112e-06, + "loss": 0.2312, + "num_input_tokens_seen": 7743016, + "step": 28890 + }, + { + "epoch": 7.574049803407601, + "grad_norm": 1.1605089902877808, + "learning_rate": 8.442423844187647e-06, + "loss": 0.1186, + "num_input_tokens_seen": 7743928, + "step": 28895 + }, + { + "epoch": 7.575360419397117, + "grad_norm": 1.5523402690887451, + "learning_rate": 8.433856353017261e-06, + "loss": 0.1307, + "num_input_tokens_seen": 7745320, + "step": 28900 + }, + { + "epoch": 7.576671035386632, + "grad_norm": 0.9152292013168335, + "learning_rate": 8.425292329112124e-06, + "loss": 0.1158, + "num_input_tokens_seen": 7746584, + "step": 28905 + }, + { + "epoch": 7.577981651376147, + "grad_norm": 2.1975603103637695, + "learning_rate": 8.416731774264658e-06, + "loss": 0.2377, + "num_input_tokens_seen": 7748168, + "step": 28910 + }, + { + "epoch": 7.579292267365662, + "grad_norm": 1.3562543392181396, + "learning_rate": 8.408174690266577e-06, + "loss": 0.1742, + "num_input_tokens_seen": 7749400, + "step": 28915 + }, + { + "epoch": 7.580602883355177, + "grad_norm": 1.9593685865402222, + "learning_rate": 8.399621078908865e-06, + "loss": 0.1367, + "num_input_tokens_seen": 7750488, + "step": 28920 + }, + { + "epoch": 7.581913499344692, + "grad_norm": 3.351236343383789, + "learning_rate": 8.391070941981778e-06, + "loss": 0.2637, + "num_input_tokens_seen": 7751512, + "step": 28925 + }, + { + "epoch": 7.5832241153342075, + "grad_norm": 1.8377625942230225, + "learning_rate": 8.382524281274844e-06, + "loss": 0.2389, + "num_input_tokens_seen": 7752504, + "step": 28930 + }, + { + "epoch": 7.584534731323722, + "grad_norm": 1.1689649820327759, + "learning_rate": 8.373981098576865e-06, + "loss": 0.2442, + "num_input_tokens_seen": 7754120, + "step": 28935 + }, + { + "epoch": 7.585845347313237, + "grad_norm": 2.6669743061065674, + "learning_rate": 8.365441395675907e-06, + "loss": 0.2255, + "num_input_tokens_seen": 7755640, + "step": 28940 + }, + { + "epoch": 7.587155963302752, + "grad_norm": 3.224820852279663, + "learning_rate": 8.356905174359338e-06, + "loss": 0.1553, + "num_input_tokens_seen": 7757048, + "step": 28945 + }, + { + "epoch": 7.588466579292267, + "grad_norm": 3.3236277103424072, + "learning_rate": 8.348372436413754e-06, + "loss": 0.1661, + "num_input_tokens_seen": 7758232, + "step": 28950 + }, + { + "epoch": 7.5897771952817825, + "grad_norm": 1.9551483392715454, + "learning_rate": 8.339843183625046e-06, + "loss": 0.1877, + "num_input_tokens_seen": 7759576, + "step": 28955 + }, + { + "epoch": 7.591087811271297, + "grad_norm": 0.658724844455719, + "learning_rate": 8.33131741777837e-06, + "loss": 0.1838, + "num_input_tokens_seen": 7761032, + "step": 28960 + }, + { + "epoch": 7.592398427260813, + "grad_norm": 2.858755350112915, + "learning_rate": 8.32279514065816e-06, + "loss": 0.1707, + "num_input_tokens_seen": 7762392, + "step": 28965 + }, + { + "epoch": 7.593709043250327, + "grad_norm": 3.586280107498169, + "learning_rate": 8.314276354048109e-06, + "loss": 0.2644, + "num_input_tokens_seen": 7763992, + "step": 28970 + }, + { + "epoch": 7.595019659239843, + "grad_norm": 1.5200440883636475, + "learning_rate": 8.30576105973119e-06, + "loss": 0.2363, + "num_input_tokens_seen": 7764936, + "step": 28975 + }, + { + "epoch": 7.5963302752293576, + "grad_norm": 1.862017273902893, + "learning_rate": 8.297249259489634e-06, + "loss": 0.2531, + "num_input_tokens_seen": 7766728, + "step": 28980 + }, + { + "epoch": 7.597640891218873, + "grad_norm": 1.820535659790039, + "learning_rate": 8.288740955104946e-06, + "loss": 0.1444, + "num_input_tokens_seen": 7768088, + "step": 28985 + }, + { + "epoch": 7.598951507208388, + "grad_norm": 1.5195753574371338, + "learning_rate": 8.280236148357911e-06, + "loss": 0.1922, + "num_input_tokens_seen": 7769416, + "step": 28990 + }, + { + "epoch": 7.600262123197903, + "grad_norm": 0.9268987774848938, + "learning_rate": 8.271734841028553e-06, + "loss": 0.1441, + "num_input_tokens_seen": 7770552, + "step": 28995 + }, + { + "epoch": 7.601572739187418, + "grad_norm": 1.5978132486343384, + "learning_rate": 8.263237034896179e-06, + "loss": 0.1716, + "num_input_tokens_seen": 7772232, + "step": 29000 + }, + { + "epoch": 7.6028833551769335, + "grad_norm": 1.696695327758789, + "learning_rate": 8.254742731739384e-06, + "loss": 0.1892, + "num_input_tokens_seen": 7773432, + "step": 29005 + }, + { + "epoch": 7.604193971166448, + "grad_norm": 2.3612351417541504, + "learning_rate": 8.246251933336e-06, + "loss": 0.1786, + "num_input_tokens_seen": 7774472, + "step": 29010 + }, + { + "epoch": 7.605504587155964, + "grad_norm": 0.651604950428009, + "learning_rate": 8.237764641463139e-06, + "loss": 0.1191, + "num_input_tokens_seen": 7775448, + "step": 29015 + }, + { + "epoch": 7.606815203145478, + "grad_norm": 1.0049183368682861, + "learning_rate": 8.229280857897176e-06, + "loss": 0.2219, + "num_input_tokens_seen": 7776520, + "step": 29020 + }, + { + "epoch": 7.608125819134994, + "grad_norm": 2.386051654815674, + "learning_rate": 8.22080058441376e-06, + "loss": 0.3057, + "num_input_tokens_seen": 7778200, + "step": 29025 + }, + { + "epoch": 7.6094364351245085, + "grad_norm": 1.8840208053588867, + "learning_rate": 8.212323822787782e-06, + "loss": 0.2024, + "num_input_tokens_seen": 7779432, + "step": 29030 + }, + { + "epoch": 7.610747051114023, + "grad_norm": 1.7357583045959473, + "learning_rate": 8.20385057479342e-06, + "loss": 0.2262, + "num_input_tokens_seen": 7781032, + "step": 29035 + }, + { + "epoch": 7.612057667103539, + "grad_norm": 1.932982325553894, + "learning_rate": 8.195380842204112e-06, + "loss": 0.2644, + "num_input_tokens_seen": 7782936, + "step": 29040 + }, + { + "epoch": 7.613368283093053, + "grad_norm": 1.204574704170227, + "learning_rate": 8.186914626792558e-06, + "loss": 0.1689, + "num_input_tokens_seen": 7784280, + "step": 29045 + }, + { + "epoch": 7.614678899082569, + "grad_norm": 3.0758309364318848, + "learning_rate": 8.178451930330719e-06, + "loss": 0.2334, + "num_input_tokens_seen": 7785688, + "step": 29050 + }, + { + "epoch": 7.6159895150720835, + "grad_norm": 2.7886157035827637, + "learning_rate": 8.169992754589819e-06, + "loss": 0.36, + "num_input_tokens_seen": 7786952, + "step": 29055 + }, + { + "epoch": 7.617300131061599, + "grad_norm": 5.390568256378174, + "learning_rate": 8.161537101340375e-06, + "loss": 0.2597, + "num_input_tokens_seen": 7787864, + "step": 29060 + }, + { + "epoch": 7.618610747051114, + "grad_norm": 2.515324354171753, + "learning_rate": 8.15308497235211e-06, + "loss": 0.2185, + "num_input_tokens_seen": 7789320, + "step": 29065 + }, + { + "epoch": 7.619921363040629, + "grad_norm": 5.207927703857422, + "learning_rate": 8.144636369394048e-06, + "loss": 0.2264, + "num_input_tokens_seen": 7790680, + "step": 29070 + }, + { + "epoch": 7.621231979030144, + "grad_norm": 4.680813312530518, + "learning_rate": 8.136191294234474e-06, + "loss": 0.1938, + "num_input_tokens_seen": 7791976, + "step": 29075 + }, + { + "epoch": 7.6225425950196595, + "grad_norm": 2.3999569416046143, + "learning_rate": 8.127749748640923e-06, + "loss": 0.2082, + "num_input_tokens_seen": 7793256, + "step": 29080 + }, + { + "epoch": 7.623853211009174, + "grad_norm": 1.6997294425964355, + "learning_rate": 8.119311734380197e-06, + "loss": 0.1307, + "num_input_tokens_seen": 7794408, + "step": 29085 + }, + { + "epoch": 7.62516382699869, + "grad_norm": 2.8567183017730713, + "learning_rate": 8.110877253218358e-06, + "loss": 0.2023, + "num_input_tokens_seen": 7795608, + "step": 29090 + }, + { + "epoch": 7.626474442988204, + "grad_norm": 5.550291538238525, + "learning_rate": 8.102446306920725e-06, + "loss": 0.2908, + "num_input_tokens_seen": 7796968, + "step": 29095 + }, + { + "epoch": 7.62778505897772, + "grad_norm": 1.9899301528930664, + "learning_rate": 8.094018897251887e-06, + "loss": 0.1881, + "num_input_tokens_seen": 7798088, + "step": 29100 + }, + { + "epoch": 7.6290956749672345, + "grad_norm": 1.7458869218826294, + "learning_rate": 8.085595025975687e-06, + "loss": 0.2073, + "num_input_tokens_seen": 7799272, + "step": 29105 + }, + { + "epoch": 7.63040629095675, + "grad_norm": 3.367037296295166, + "learning_rate": 8.077174694855216e-06, + "loss": 0.2214, + "num_input_tokens_seen": 7800648, + "step": 29110 + }, + { + "epoch": 7.631716906946265, + "grad_norm": 3.845050096511841, + "learning_rate": 8.068757905652832e-06, + "loss": 0.1661, + "num_input_tokens_seen": 7802200, + "step": 29115 + }, + { + "epoch": 7.63302752293578, + "grad_norm": 1.7968535423278809, + "learning_rate": 8.060344660130172e-06, + "loss": 0.2171, + "num_input_tokens_seen": 7803304, + "step": 29120 + }, + { + "epoch": 7.634338138925295, + "grad_norm": 1.8009322881698608, + "learning_rate": 8.051934960048104e-06, + "loss": 0.1575, + "num_input_tokens_seen": 7804136, + "step": 29125 + }, + { + "epoch": 7.6356487549148095, + "grad_norm": 3.395371913909912, + "learning_rate": 8.043528807166767e-06, + "loss": 0.3155, + "num_input_tokens_seen": 7805608, + "step": 29130 + }, + { + "epoch": 7.636959370904325, + "grad_norm": 1.876670002937317, + "learning_rate": 8.035126203245552e-06, + "loss": 0.1423, + "num_input_tokens_seen": 7806808, + "step": 29135 + }, + { + "epoch": 7.63826998689384, + "grad_norm": 2.093515396118164, + "learning_rate": 8.026727150043118e-06, + "loss": 0.238, + "num_input_tokens_seen": 7808312, + "step": 29140 + }, + { + "epoch": 7.639580602883355, + "grad_norm": 3.165501356124878, + "learning_rate": 8.018331649317353e-06, + "loss": 0.1327, + "num_input_tokens_seen": 7809496, + "step": 29145 + }, + { + "epoch": 7.64089121887287, + "grad_norm": 1.4587138891220093, + "learning_rate": 8.009939702825433e-06, + "loss": 0.2315, + "num_input_tokens_seen": 7810792, + "step": 29150 + }, + { + "epoch": 7.6422018348623855, + "grad_norm": 1.1587305068969727, + "learning_rate": 8.001551312323776e-06, + "loss": 0.158, + "num_input_tokens_seen": 7811832, + "step": 29155 + }, + { + "epoch": 7.6435124508519, + "grad_norm": 2.200530529022217, + "learning_rate": 7.993166479568063e-06, + "loss": 0.1487, + "num_input_tokens_seen": 7813032, + "step": 29160 + }, + { + "epoch": 7.644823066841416, + "grad_norm": 2.384004831314087, + "learning_rate": 7.984785206313213e-06, + "loss": 0.2398, + "num_input_tokens_seen": 7814472, + "step": 29165 + }, + { + "epoch": 7.64613368283093, + "grad_norm": 3.7048745155334473, + "learning_rate": 7.976407494313418e-06, + "loss": 0.2023, + "num_input_tokens_seen": 7816312, + "step": 29170 + }, + { + "epoch": 7.647444298820446, + "grad_norm": 1.2690256834030151, + "learning_rate": 7.968033345322123e-06, + "loss": 0.1471, + "num_input_tokens_seen": 7817736, + "step": 29175 + }, + { + "epoch": 7.6487549148099605, + "grad_norm": 2.4327950477600098, + "learning_rate": 7.95966276109203e-06, + "loss": 0.1396, + "num_input_tokens_seen": 7818984, + "step": 29180 + }, + { + "epoch": 7.650065530799476, + "grad_norm": 1.3825534582138062, + "learning_rate": 7.95129574337507e-06, + "loss": 0.2793, + "num_input_tokens_seen": 7820280, + "step": 29185 + }, + { + "epoch": 7.651376146788991, + "grad_norm": 1.746997356414795, + "learning_rate": 7.942932293922453e-06, + "loss": 0.1857, + "num_input_tokens_seen": 7821544, + "step": 29190 + }, + { + "epoch": 7.652686762778506, + "grad_norm": 0.9271208047866821, + "learning_rate": 7.934572414484634e-06, + "loss": 0.1434, + "num_input_tokens_seen": 7822776, + "step": 29195 + }, + { + "epoch": 7.653997378768021, + "grad_norm": 2.377573013305664, + "learning_rate": 7.926216106811321e-06, + "loss": 0.1456, + "num_input_tokens_seen": 7824008, + "step": 29200 + }, + { + "epoch": 7.655307994757536, + "grad_norm": 1.4481793642044067, + "learning_rate": 7.917863372651477e-06, + "loss": 0.214, + "num_input_tokens_seen": 7825528, + "step": 29205 + }, + { + "epoch": 7.656618610747051, + "grad_norm": 1.4878731966018677, + "learning_rate": 7.909514213753314e-06, + "loss": 0.1285, + "num_input_tokens_seen": 7826952, + "step": 29210 + }, + { + "epoch": 7.657929226736567, + "grad_norm": 1.1286044120788574, + "learning_rate": 7.901168631864294e-06, + "loss": 0.1871, + "num_input_tokens_seen": 7828152, + "step": 29215 + }, + { + "epoch": 7.659239842726081, + "grad_norm": 1.4950320720672607, + "learning_rate": 7.892826628731135e-06, + "loss": 0.3198, + "num_input_tokens_seen": 7829864, + "step": 29220 + }, + { + "epoch": 7.660550458715596, + "grad_norm": 4.547328472137451, + "learning_rate": 7.884488206099814e-06, + "loss": 0.1838, + "num_input_tokens_seen": 7831096, + "step": 29225 + }, + { + "epoch": 7.6618610747051115, + "grad_norm": 2.1820497512817383, + "learning_rate": 7.876153365715519e-06, + "loss": 0.2283, + "num_input_tokens_seen": 7832392, + "step": 29230 + }, + { + "epoch": 7.663171690694626, + "grad_norm": 1.932407259941101, + "learning_rate": 7.867822109322747e-06, + "loss": 0.1794, + "num_input_tokens_seen": 7833528, + "step": 29235 + }, + { + "epoch": 7.664482306684142, + "grad_norm": 2.5439248085021973, + "learning_rate": 7.859494438665204e-06, + "loss": 0.2338, + "num_input_tokens_seen": 7835064, + "step": 29240 + }, + { + "epoch": 7.665792922673656, + "grad_norm": 2.3149685859680176, + "learning_rate": 7.85117035548586e-06, + "loss": 0.1282, + "num_input_tokens_seen": 7836440, + "step": 29245 + }, + { + "epoch": 7.667103538663172, + "grad_norm": 5.506032943725586, + "learning_rate": 7.842849861526929e-06, + "loss": 0.2502, + "num_input_tokens_seen": 7837736, + "step": 29250 + }, + { + "epoch": 7.6684141546526865, + "grad_norm": 3.7589614391326904, + "learning_rate": 7.834532958529878e-06, + "loss": 0.1531, + "num_input_tokens_seen": 7838920, + "step": 29255 + }, + { + "epoch": 7.669724770642202, + "grad_norm": 2.026242733001709, + "learning_rate": 7.826219648235424e-06, + "loss": 0.1391, + "num_input_tokens_seen": 7840296, + "step": 29260 + }, + { + "epoch": 7.671035386631717, + "grad_norm": 2.5944273471832275, + "learning_rate": 7.817909932383519e-06, + "loss": 0.3512, + "num_input_tokens_seen": 7841928, + "step": 29265 + }, + { + "epoch": 7.672346002621232, + "grad_norm": 2.771648406982422, + "learning_rate": 7.809603812713373e-06, + "loss": 0.1721, + "num_input_tokens_seen": 7843000, + "step": 29270 + }, + { + "epoch": 7.673656618610747, + "grad_norm": 1.5001963376998901, + "learning_rate": 7.80130129096345e-06, + "loss": 0.2855, + "num_input_tokens_seen": 7844616, + "step": 29275 + }, + { + "epoch": 7.674967234600262, + "grad_norm": 6.743342876434326, + "learning_rate": 7.793002368871446e-06, + "loss": 0.1903, + "num_input_tokens_seen": 7845752, + "step": 29280 + }, + { + "epoch": 7.676277850589777, + "grad_norm": 1.9937478303909302, + "learning_rate": 7.78470704817431e-06, + "loss": 0.1655, + "num_input_tokens_seen": 7847112, + "step": 29285 + }, + { + "epoch": 7.677588466579293, + "grad_norm": 3.16805100440979, + "learning_rate": 7.77641533060825e-06, + "loss": 0.1413, + "num_input_tokens_seen": 7848264, + "step": 29290 + }, + { + "epoch": 7.678899082568807, + "grad_norm": 1.8616509437561035, + "learning_rate": 7.76812721790871e-06, + "loss": 0.2575, + "num_input_tokens_seen": 7849672, + "step": 29295 + }, + { + "epoch": 7.680209698558322, + "grad_norm": 2.495698928833008, + "learning_rate": 7.759842711810361e-06, + "loss": 0.2388, + "num_input_tokens_seen": 7850952, + "step": 29300 + }, + { + "epoch": 7.6815203145478375, + "grad_norm": 2.2472808361053467, + "learning_rate": 7.751561814047142e-06, + "loss": 0.3193, + "num_input_tokens_seen": 7852216, + "step": 29305 + }, + { + "epoch": 7.682830930537353, + "grad_norm": 1.357776403427124, + "learning_rate": 7.743284526352235e-06, + "loss": 0.1397, + "num_input_tokens_seen": 7853368, + "step": 29310 + }, + { + "epoch": 7.684141546526868, + "grad_norm": 2.7087199687957764, + "learning_rate": 7.735010850458057e-06, + "loss": 0.2549, + "num_input_tokens_seen": 7855080, + "step": 29315 + }, + { + "epoch": 7.685452162516382, + "grad_norm": 2.7374048233032227, + "learning_rate": 7.726740788096276e-06, + "loss": 0.2116, + "num_input_tokens_seen": 7856536, + "step": 29320 + }, + { + "epoch": 7.686762778505898, + "grad_norm": 2.8108818531036377, + "learning_rate": 7.718474340997804e-06, + "loss": 0.3091, + "num_input_tokens_seen": 7858104, + "step": 29325 + }, + { + "epoch": 7.6880733944954125, + "grad_norm": 2.97733998298645, + "learning_rate": 7.710211510892792e-06, + "loss": 0.3406, + "num_input_tokens_seen": 7859752, + "step": 29330 + }, + { + "epoch": 7.689384010484928, + "grad_norm": 2.408170223236084, + "learning_rate": 7.701952299510634e-06, + "loss": 0.1943, + "num_input_tokens_seen": 7862760, + "step": 29335 + }, + { + "epoch": 7.690694626474443, + "grad_norm": 1.8685897588729858, + "learning_rate": 7.69369670857998e-06, + "loss": 0.1738, + "num_input_tokens_seen": 7864632, + "step": 29340 + }, + { + "epoch": 7.692005242463958, + "grad_norm": 1.7952783107757568, + "learning_rate": 7.685444739828688e-06, + "loss": 0.1281, + "num_input_tokens_seen": 7866152, + "step": 29345 + }, + { + "epoch": 7.693315858453473, + "grad_norm": 3.8639018535614014, + "learning_rate": 7.677196394983902e-06, + "loss": 0.1849, + "num_input_tokens_seen": 7867336, + "step": 29350 + }, + { + "epoch": 7.694626474442988, + "grad_norm": 4.092068195343018, + "learning_rate": 7.668951675771979e-06, + "loss": 0.1859, + "num_input_tokens_seen": 7868568, + "step": 29355 + }, + { + "epoch": 7.695937090432503, + "grad_norm": 2.410219669342041, + "learning_rate": 7.660710583918524e-06, + "loss": 0.1822, + "num_input_tokens_seen": 7869784, + "step": 29360 + }, + { + "epoch": 7.697247706422019, + "grad_norm": 1.5188205242156982, + "learning_rate": 7.652473121148383e-06, + "loss": 0.3343, + "num_input_tokens_seen": 7871112, + "step": 29365 + }, + { + "epoch": 7.698558322411533, + "grad_norm": 3.28926157951355, + "learning_rate": 7.644239289185644e-06, + "loss": 0.2481, + "num_input_tokens_seen": 7872776, + "step": 29370 + }, + { + "epoch": 7.699868938401049, + "grad_norm": 2.0586085319519043, + "learning_rate": 7.636009089753643e-06, + "loss": 0.1546, + "num_input_tokens_seen": 7874056, + "step": 29375 + }, + { + "epoch": 7.7011795543905635, + "grad_norm": 3.489563226699829, + "learning_rate": 7.627782524574925e-06, + "loss": 0.2443, + "num_input_tokens_seen": 7875224, + "step": 29380 + }, + { + "epoch": 7.702490170380079, + "grad_norm": 1.0490199327468872, + "learning_rate": 7.619559595371312e-06, + "loss": 0.1775, + "num_input_tokens_seen": 7876536, + "step": 29385 + }, + { + "epoch": 7.703800786369594, + "grad_norm": 2.2011990547180176, + "learning_rate": 7.611340303863843e-06, + "loss": 0.171, + "num_input_tokens_seen": 7877688, + "step": 29390 + }, + { + "epoch": 7.705111402359108, + "grad_norm": 1.7286076545715332, + "learning_rate": 7.603124651772803e-06, + "loss": 0.2037, + "num_input_tokens_seen": 7878968, + "step": 29395 + }, + { + "epoch": 7.706422018348624, + "grad_norm": 1.1096687316894531, + "learning_rate": 7.594912640817705e-06, + "loss": 0.0796, + "num_input_tokens_seen": 7880312, + "step": 29400 + }, + { + "epoch": 7.707732634338139, + "grad_norm": 1.173767328262329, + "learning_rate": 7.586704272717327e-06, + "loss": 0.2009, + "num_input_tokens_seen": 7881832, + "step": 29405 + }, + { + "epoch": 7.709043250327654, + "grad_norm": 2.179387092590332, + "learning_rate": 7.578499549189655e-06, + "loss": 0.1215, + "num_input_tokens_seen": 7882984, + "step": 29410 + }, + { + "epoch": 7.710353866317169, + "grad_norm": 1.5779825448989868, + "learning_rate": 7.570298471951934e-06, + "loss": 0.1318, + "num_input_tokens_seen": 7884440, + "step": 29415 + }, + { + "epoch": 7.711664482306684, + "grad_norm": 1.3528555631637573, + "learning_rate": 7.562101042720621e-06, + "loss": 0.3664, + "num_input_tokens_seen": 7885800, + "step": 29420 + }, + { + "epoch": 7.712975098296199, + "grad_norm": 2.760148286819458, + "learning_rate": 7.553907263211427e-06, + "loss": 0.2453, + "num_input_tokens_seen": 7886968, + "step": 29425 + }, + { + "epoch": 7.714285714285714, + "grad_norm": 1.339493989944458, + "learning_rate": 7.545717135139299e-06, + "loss": 0.188, + "num_input_tokens_seen": 7888904, + "step": 29430 + }, + { + "epoch": 7.715596330275229, + "grad_norm": 1.8804020881652832, + "learning_rate": 7.537530660218414e-06, + "loss": 0.1203, + "num_input_tokens_seen": 7889944, + "step": 29435 + }, + { + "epoch": 7.716906946264745, + "grad_norm": 1.8882328271865845, + "learning_rate": 7.529347840162193e-06, + "loss": 0.2101, + "num_input_tokens_seen": 7891048, + "step": 29440 + }, + { + "epoch": 7.718217562254259, + "grad_norm": 5.69964599609375, + "learning_rate": 7.52116867668328e-06, + "loss": 0.2084, + "num_input_tokens_seen": 7892504, + "step": 29445 + }, + { + "epoch": 7.719528178243775, + "grad_norm": 0.7151444554328918, + "learning_rate": 7.51299317149356e-06, + "loss": 0.2564, + "num_input_tokens_seen": 7893848, + "step": 29450 + }, + { + "epoch": 7.7208387942332894, + "grad_norm": 2.135981321334839, + "learning_rate": 7.504821326304157e-06, + "loss": 0.1654, + "num_input_tokens_seen": 7895192, + "step": 29455 + }, + { + "epoch": 7.722149410222805, + "grad_norm": 3.445596694946289, + "learning_rate": 7.496653142825419e-06, + "loss": 0.1511, + "num_input_tokens_seen": 7896424, + "step": 29460 + }, + { + "epoch": 7.72346002621232, + "grad_norm": 4.725684642791748, + "learning_rate": 7.4884886227669375e-06, + "loss": 0.1353, + "num_input_tokens_seen": 7897992, + "step": 29465 + }, + { + "epoch": 7.724770642201835, + "grad_norm": 3.4395437240600586, + "learning_rate": 7.480327767837525e-06, + "loss": 0.1758, + "num_input_tokens_seen": 7899128, + "step": 29470 + }, + { + "epoch": 7.72608125819135, + "grad_norm": 0.277937650680542, + "learning_rate": 7.472170579745244e-06, + "loss": 0.2879, + "num_input_tokens_seen": 7900568, + "step": 29475 + }, + { + "epoch": 7.727391874180865, + "grad_norm": 5.779224395751953, + "learning_rate": 7.464017060197373e-06, + "loss": 0.214, + "num_input_tokens_seen": 7902024, + "step": 29480 + }, + { + "epoch": 7.72870249017038, + "grad_norm": 1.4498482942581177, + "learning_rate": 7.45586721090043e-06, + "loss": 0.0902, + "num_input_tokens_seen": 7903416, + "step": 29485 + }, + { + "epoch": 7.730013106159895, + "grad_norm": 1.1104662418365479, + "learning_rate": 7.447721033560168e-06, + "loss": 0.202, + "num_input_tokens_seen": 7904552, + "step": 29490 + }, + { + "epoch": 7.73132372214941, + "grad_norm": 1.3252904415130615, + "learning_rate": 7.439578529881572e-06, + "loss": 0.146, + "num_input_tokens_seen": 7905752, + "step": 29495 + }, + { + "epoch": 7.732634338138926, + "grad_norm": 2.4823827743530273, + "learning_rate": 7.431439701568841e-06, + "loss": 0.1485, + "num_input_tokens_seen": 7907240, + "step": 29500 + }, + { + "epoch": 7.73394495412844, + "grad_norm": 1.9497076272964478, + "learning_rate": 7.423304550325422e-06, + "loss": 0.2169, + "num_input_tokens_seen": 7908296, + "step": 29505 + }, + { + "epoch": 7.735255570117955, + "grad_norm": 2.2250373363494873, + "learning_rate": 7.415173077853993e-06, + "loss": 0.1363, + "num_input_tokens_seen": 7909384, + "step": 29510 + }, + { + "epoch": 7.736566186107471, + "grad_norm": 0.2740934491157532, + "learning_rate": 7.407045285856445e-06, + "loss": 0.2264, + "num_input_tokens_seen": 7911080, + "step": 29515 + }, + { + "epoch": 7.737876802096985, + "grad_norm": 1.6310940980911255, + "learning_rate": 7.39892117603393e-06, + "loss": 0.2178, + "num_input_tokens_seen": 7911976, + "step": 29520 + }, + { + "epoch": 7.739187418086501, + "grad_norm": 1.3686312437057495, + "learning_rate": 7.390800750086799e-06, + "loss": 0.2724, + "num_input_tokens_seen": 7913416, + "step": 29525 + }, + { + "epoch": 7.740498034076015, + "grad_norm": 3.163273334503174, + "learning_rate": 7.382684009714652e-06, + "loss": 0.2077, + "num_input_tokens_seen": 7914936, + "step": 29530 + }, + { + "epoch": 7.741808650065531, + "grad_norm": 5.009366035461426, + "learning_rate": 7.374570956616295e-06, + "loss": 0.4136, + "num_input_tokens_seen": 7918200, + "step": 29535 + }, + { + "epoch": 7.743119266055046, + "grad_norm": 0.06171748414635658, + "learning_rate": 7.366461592489782e-06, + "loss": 0.2349, + "num_input_tokens_seen": 7919208, + "step": 29540 + }, + { + "epoch": 7.744429882044561, + "grad_norm": 1.1172603368759155, + "learning_rate": 7.358355919032389e-06, + "loss": 0.1775, + "num_input_tokens_seen": 7920648, + "step": 29545 + }, + { + "epoch": 7.745740498034076, + "grad_norm": 3.8855972290039062, + "learning_rate": 7.350253937940621e-06, + "loss": 0.123, + "num_input_tokens_seen": 7921544, + "step": 29550 + }, + { + "epoch": 7.747051114023591, + "grad_norm": 1.9101669788360596, + "learning_rate": 7.342155650910207e-06, + "loss": 0.1088, + "num_input_tokens_seen": 7922648, + "step": 29555 + }, + { + "epoch": 7.748361730013106, + "grad_norm": 2.672630786895752, + "learning_rate": 7.334061059636102e-06, + "loss": 0.1437, + "num_input_tokens_seen": 7923880, + "step": 29560 + }, + { + "epoch": 7.749672346002622, + "grad_norm": 1.9896376132965088, + "learning_rate": 7.325970165812496e-06, + "loss": 0.1718, + "num_input_tokens_seen": 7925592, + "step": 29565 + }, + { + "epoch": 7.750982961992136, + "grad_norm": 1.4870402812957764, + "learning_rate": 7.3178829711327945e-06, + "loss": 0.1819, + "num_input_tokens_seen": 7926856, + "step": 29570 + }, + { + "epoch": 7.752293577981652, + "grad_norm": 3.7164416313171387, + "learning_rate": 7.309799477289636e-06, + "loss": 0.2214, + "num_input_tokens_seen": 7928056, + "step": 29575 + }, + { + "epoch": 7.753604193971166, + "grad_norm": 3.341090440750122, + "learning_rate": 7.301719685974881e-06, + "loss": 0.2201, + "num_input_tokens_seen": 7929032, + "step": 29580 + }, + { + "epoch": 7.754914809960681, + "grad_norm": 1.6598703861236572, + "learning_rate": 7.293643598879612e-06, + "loss": 0.1874, + "num_input_tokens_seen": 7930056, + "step": 29585 + }, + { + "epoch": 7.756225425950197, + "grad_norm": 1.5673327445983887, + "learning_rate": 7.285571217694148e-06, + "loss": 0.2376, + "num_input_tokens_seen": 7931288, + "step": 29590 + }, + { + "epoch": 7.757536041939712, + "grad_norm": 3.9245498180389404, + "learning_rate": 7.277502544108017e-06, + "loss": 0.2171, + "num_input_tokens_seen": 7932424, + "step": 29595 + }, + { + "epoch": 7.758846657929227, + "grad_norm": 0.6412496566772461, + "learning_rate": 7.2694375798099815e-06, + "loss": 0.1685, + "num_input_tokens_seen": 7933880, + "step": 29600 + }, + { + "epoch": 7.760157273918741, + "grad_norm": 3.9706153869628906, + "learning_rate": 7.261376326488026e-06, + "loss": 0.3557, + "num_input_tokens_seen": 7935080, + "step": 29605 + }, + { + "epoch": 7.761467889908257, + "grad_norm": 2.3104629516601562, + "learning_rate": 7.253318785829363e-06, + "loss": 0.2368, + "num_input_tokens_seen": 7936168, + "step": 29610 + }, + { + "epoch": 7.762778505897772, + "grad_norm": 1.7534946203231812, + "learning_rate": 7.245264959520406e-06, + "loss": 0.1456, + "num_input_tokens_seen": 7937352, + "step": 29615 + }, + { + "epoch": 7.764089121887287, + "grad_norm": 0.26047009229660034, + "learning_rate": 7.237214849246815e-06, + "loss": 0.1666, + "num_input_tokens_seen": 7938712, + "step": 29620 + }, + { + "epoch": 7.765399737876802, + "grad_norm": 2.6724116802215576, + "learning_rate": 7.2291684566934634e-06, + "loss": 0.3286, + "num_input_tokens_seen": 7940664, + "step": 29625 + }, + { + "epoch": 7.766710353866317, + "grad_norm": 2.289109706878662, + "learning_rate": 7.221125783544441e-06, + "loss": 0.2081, + "num_input_tokens_seen": 7942088, + "step": 29630 + }, + { + "epoch": 7.768020969855832, + "grad_norm": 1.5915429592132568, + "learning_rate": 7.213086831483082e-06, + "loss": 0.1321, + "num_input_tokens_seen": 7943352, + "step": 29635 + }, + { + "epoch": 7.7693315858453476, + "grad_norm": 3.1295833587646484, + "learning_rate": 7.205051602191914e-06, + "loss": 0.1755, + "num_input_tokens_seen": 7944728, + "step": 29640 + }, + { + "epoch": 7.770642201834862, + "grad_norm": 3.3983664512634277, + "learning_rate": 7.1970200973527e-06, + "loss": 0.2649, + "num_input_tokens_seen": 7946072, + "step": 29645 + }, + { + "epoch": 7.771952817824378, + "grad_norm": 1.798100471496582, + "learning_rate": 7.188992318646423e-06, + "loss": 0.1881, + "num_input_tokens_seen": 7946920, + "step": 29650 + }, + { + "epoch": 7.773263433813892, + "grad_norm": 2.9715869426727295, + "learning_rate": 7.180968267753274e-06, + "loss": 0.2668, + "num_input_tokens_seen": 7948616, + "step": 29655 + }, + { + "epoch": 7.774574049803408, + "grad_norm": 1.4362175464630127, + "learning_rate": 7.172947946352676e-06, + "loss": 0.1905, + "num_input_tokens_seen": 7950600, + "step": 29660 + }, + { + "epoch": 7.775884665792923, + "grad_norm": 2.7132115364074707, + "learning_rate": 7.164931356123269e-06, + "loss": 0.2299, + "num_input_tokens_seen": 7951752, + "step": 29665 + }, + { + "epoch": 7.777195281782438, + "grad_norm": 2.305661678314209, + "learning_rate": 7.156918498742915e-06, + "loss": 0.2351, + "num_input_tokens_seen": 7953256, + "step": 29670 + }, + { + "epoch": 7.778505897771953, + "grad_norm": 1.523239016532898, + "learning_rate": 7.148909375888688e-06, + "loss": 0.086, + "num_input_tokens_seen": 7954488, + "step": 29675 + }, + { + "epoch": 7.779816513761467, + "grad_norm": 1.759261131286621, + "learning_rate": 7.140903989236886e-06, + "loss": 0.2752, + "num_input_tokens_seen": 7955816, + "step": 29680 + }, + { + "epoch": 7.781127129750983, + "grad_norm": 2.794326066970825, + "learning_rate": 7.132902340463018e-06, + "loss": 0.2498, + "num_input_tokens_seen": 7957304, + "step": 29685 + }, + { + "epoch": 7.7824377457404985, + "grad_norm": 1.8906344175338745, + "learning_rate": 7.124904431241819e-06, + "loss": 0.3091, + "num_input_tokens_seen": 7958616, + "step": 29690 + }, + { + "epoch": 7.783748361730013, + "grad_norm": 1.3098224401474, + "learning_rate": 7.116910263247237e-06, + "loss": 0.1254, + "num_input_tokens_seen": 7959848, + "step": 29695 + }, + { + "epoch": 7.785058977719528, + "grad_norm": 1.527644157409668, + "learning_rate": 7.108919838152436e-06, + "loss": 0.208, + "num_input_tokens_seen": 7961048, + "step": 29700 + }, + { + "epoch": 7.786369593709043, + "grad_norm": 1.6567496061325073, + "learning_rate": 7.100933157629802e-06, + "loss": 0.2412, + "num_input_tokens_seen": 7962648, + "step": 29705 + }, + { + "epoch": 7.787680209698558, + "grad_norm": 1.033992052078247, + "learning_rate": 7.09295022335093e-06, + "loss": 0.2704, + "num_input_tokens_seen": 7964008, + "step": 29710 + }, + { + "epoch": 7.7889908256880735, + "grad_norm": 1.3981778621673584, + "learning_rate": 7.0849710369866345e-06, + "loss": 0.2455, + "num_input_tokens_seen": 7965176, + "step": 29715 + }, + { + "epoch": 7.790301441677588, + "grad_norm": 3.696157217025757, + "learning_rate": 7.076995600206948e-06, + "loss": 0.2208, + "num_input_tokens_seen": 7966568, + "step": 29720 + }, + { + "epoch": 7.791612057667104, + "grad_norm": 1.5821877717971802, + "learning_rate": 7.069023914681111e-06, + "loss": 0.2394, + "num_input_tokens_seen": 7967880, + "step": 29725 + }, + { + "epoch": 7.792922673656618, + "grad_norm": 3.433608055114746, + "learning_rate": 7.0610559820775965e-06, + "loss": 0.2955, + "num_input_tokens_seen": 7969448, + "step": 29730 + }, + { + "epoch": 7.794233289646134, + "grad_norm": 4.523678779602051, + "learning_rate": 7.05309180406406e-06, + "loss": 0.2498, + "num_input_tokens_seen": 7970488, + "step": 29735 + }, + { + "epoch": 7.795543905635649, + "grad_norm": 5.194426536560059, + "learning_rate": 7.045131382307399e-06, + "loss": 0.1972, + "num_input_tokens_seen": 7971560, + "step": 29740 + }, + { + "epoch": 7.796854521625164, + "grad_norm": 2.853562831878662, + "learning_rate": 7.0371747184737075e-06, + "loss": 0.2823, + "num_input_tokens_seen": 7973032, + "step": 29745 + }, + { + "epoch": 7.798165137614679, + "grad_norm": 1.6604859828948975, + "learning_rate": 7.02922181422832e-06, + "loss": 0.3024, + "num_input_tokens_seen": 7975176, + "step": 29750 + }, + { + "epoch": 7.799475753604194, + "grad_norm": 2.4529948234558105, + "learning_rate": 7.021272671235754e-06, + "loss": 0.3936, + "num_input_tokens_seen": 7976392, + "step": 29755 + }, + { + "epoch": 7.800786369593709, + "grad_norm": 1.5316804647445679, + "learning_rate": 7.013327291159755e-06, + "loss": 0.2025, + "num_input_tokens_seen": 7977848, + "step": 29760 + }, + { + "epoch": 7.8020969855832245, + "grad_norm": 1.5068235397338867, + "learning_rate": 7.0053856756632815e-06, + "loss": 0.1868, + "num_input_tokens_seen": 7979368, + "step": 29765 + }, + { + "epoch": 7.803407601572739, + "grad_norm": 2.3856191635131836, + "learning_rate": 6.997447826408485e-06, + "loss": 0.3531, + "num_input_tokens_seen": 7980616, + "step": 29770 + }, + { + "epoch": 7.804718217562254, + "grad_norm": 4.020510673522949, + "learning_rate": 6.989513745056753e-06, + "loss": 0.1942, + "num_input_tokens_seen": 7981688, + "step": 29775 + }, + { + "epoch": 7.806028833551769, + "grad_norm": 3.780108690261841, + "learning_rate": 6.981583433268676e-06, + "loss": 0.2228, + "num_input_tokens_seen": 7983112, + "step": 29780 + }, + { + "epoch": 7.807339449541285, + "grad_norm": 0.49290144443511963, + "learning_rate": 6.97365689270405e-06, + "loss": 0.1075, + "num_input_tokens_seen": 7984088, + "step": 29785 + }, + { + "epoch": 7.8086500655307995, + "grad_norm": 2.527768135070801, + "learning_rate": 6.965734125021892e-06, + "loss": 0.1692, + "num_input_tokens_seen": 7985512, + "step": 29790 + }, + { + "epoch": 7.809960681520314, + "grad_norm": 2.303056478500366, + "learning_rate": 6.957815131880421e-06, + "loss": 0.2548, + "num_input_tokens_seen": 7987096, + "step": 29795 + }, + { + "epoch": 7.81127129750983, + "grad_norm": 2.4976534843444824, + "learning_rate": 6.949899914937066e-06, + "loss": 0.2288, + "num_input_tokens_seen": 7988824, + "step": 29800 + }, + { + "epoch": 7.812581913499344, + "grad_norm": 2.2143800258636475, + "learning_rate": 6.941988475848474e-06, + "loss": 0.1631, + "num_input_tokens_seen": 7990408, + "step": 29805 + }, + { + "epoch": 7.81389252948886, + "grad_norm": 1.1796002388000488, + "learning_rate": 6.934080816270488e-06, + "loss": 0.2424, + "num_input_tokens_seen": 7991544, + "step": 29810 + }, + { + "epoch": 7.815203145478375, + "grad_norm": 1.4917410612106323, + "learning_rate": 6.926176937858175e-06, + "loss": 0.3012, + "num_input_tokens_seen": 7992808, + "step": 29815 + }, + { + "epoch": 7.81651376146789, + "grad_norm": 0.24769563972949982, + "learning_rate": 6.9182768422658e-06, + "loss": 0.1259, + "num_input_tokens_seen": 7994088, + "step": 29820 + }, + { + "epoch": 7.817824377457405, + "grad_norm": 1.6404266357421875, + "learning_rate": 6.910380531146837e-06, + "loss": 0.4491, + "num_input_tokens_seen": 7995496, + "step": 29825 + }, + { + "epoch": 7.81913499344692, + "grad_norm": 1.018951654434204, + "learning_rate": 6.902488006153973e-06, + "loss": 0.3341, + "num_input_tokens_seen": 7996856, + "step": 29830 + }, + { + "epoch": 7.820445609436435, + "grad_norm": 3.8410372734069824, + "learning_rate": 6.894599268939101e-06, + "loss": 0.2747, + "num_input_tokens_seen": 7998008, + "step": 29835 + }, + { + "epoch": 7.8217562254259505, + "grad_norm": 0.8917456269264221, + "learning_rate": 6.886714321153315e-06, + "loss": 0.2795, + "num_input_tokens_seen": 7999432, + "step": 29840 + }, + { + "epoch": 7.823066841415465, + "grad_norm": 1.3975108861923218, + "learning_rate": 6.8788331644469265e-06, + "loss": 0.1314, + "num_input_tokens_seen": 8000680, + "step": 29845 + }, + { + "epoch": 7.824377457404981, + "grad_norm": 0.3112669289112091, + "learning_rate": 6.870955800469453e-06, + "loss": 0.1013, + "num_input_tokens_seen": 8001880, + "step": 29850 + }, + { + "epoch": 7.825688073394495, + "grad_norm": 2.935206651687622, + "learning_rate": 6.863082230869597e-06, + "loss": 0.206, + "num_input_tokens_seen": 8003288, + "step": 29855 + }, + { + "epoch": 7.826998689384011, + "grad_norm": 1.5321571826934814, + "learning_rate": 6.855212457295285e-06, + "loss": 0.256, + "num_input_tokens_seen": 8004888, + "step": 29860 + }, + { + "epoch": 7.8283093053735255, + "grad_norm": 2.2690541744232178, + "learning_rate": 6.847346481393663e-06, + "loss": 0.1669, + "num_input_tokens_seen": 8005944, + "step": 29865 + }, + { + "epoch": 7.82961992136304, + "grad_norm": 5.139392852783203, + "learning_rate": 6.839484304811055e-06, + "loss": 0.2208, + "num_input_tokens_seen": 8007240, + "step": 29870 + }, + { + "epoch": 7.830930537352556, + "grad_norm": 1.6088355779647827, + "learning_rate": 6.831625929193006e-06, + "loss": 0.1929, + "num_input_tokens_seen": 8008440, + "step": 29875 + }, + { + "epoch": 7.832241153342071, + "grad_norm": 2.421903371810913, + "learning_rate": 6.8237713561842556e-06, + "loss": 0.196, + "num_input_tokens_seen": 8009576, + "step": 29880 + }, + { + "epoch": 7.833551769331586, + "grad_norm": 3.15983247756958, + "learning_rate": 6.815920587428759e-06, + "loss": 0.2311, + "num_input_tokens_seen": 8011288, + "step": 29885 + }, + { + "epoch": 7.834862385321101, + "grad_norm": 2.562678098678589, + "learning_rate": 6.808073624569661e-06, + "loss": 0.2279, + "num_input_tokens_seen": 8014760, + "step": 29890 + }, + { + "epoch": 7.836173001310616, + "grad_norm": 4.096179962158203, + "learning_rate": 6.8002304692493165e-06, + "loss": 0.329, + "num_input_tokens_seen": 8015736, + "step": 29895 + }, + { + "epoch": 7.837483617300131, + "grad_norm": 0.9014726281166077, + "learning_rate": 6.79239112310929e-06, + "loss": 0.1806, + "num_input_tokens_seen": 8017032, + "step": 29900 + }, + { + "epoch": 7.838794233289646, + "grad_norm": 1.1701678037643433, + "learning_rate": 6.784555587790342e-06, + "loss": 0.1839, + "num_input_tokens_seen": 8018600, + "step": 29905 + }, + { + "epoch": 7.840104849279161, + "grad_norm": 1.2098649740219116, + "learning_rate": 6.7767238649324325e-06, + "loss": 0.2603, + "num_input_tokens_seen": 8019944, + "step": 29910 + }, + { + "epoch": 7.8414154652686765, + "grad_norm": 1.8573272228240967, + "learning_rate": 6.768895956174726e-06, + "loss": 0.2834, + "num_input_tokens_seen": 8021784, + "step": 29915 + }, + { + "epoch": 7.842726081258191, + "grad_norm": 0.7833120822906494, + "learning_rate": 6.761071863155607e-06, + "loss": 0.2841, + "num_input_tokens_seen": 8022920, + "step": 29920 + }, + { + "epoch": 7.844036697247707, + "grad_norm": 1.6521981954574585, + "learning_rate": 6.7532515875126295e-06, + "loss": 0.2864, + "num_input_tokens_seen": 8024200, + "step": 29925 + }, + { + "epoch": 7.845347313237221, + "grad_norm": 3.3202130794525146, + "learning_rate": 6.7454351308825634e-06, + "loss": 0.2904, + "num_input_tokens_seen": 8025640, + "step": 29930 + }, + { + "epoch": 7.846657929226737, + "grad_norm": 2.610305070877075, + "learning_rate": 6.737622494901383e-06, + "loss": 0.3199, + "num_input_tokens_seen": 8026696, + "step": 29935 + }, + { + "epoch": 7.8479685452162515, + "grad_norm": 1.7791080474853516, + "learning_rate": 6.7298136812042615e-06, + "loss": 0.1805, + "num_input_tokens_seen": 8027768, + "step": 29940 + }, + { + "epoch": 7.849279161205767, + "grad_norm": 4.145167827606201, + "learning_rate": 6.722008691425566e-06, + "loss": 0.2143, + "num_input_tokens_seen": 8029384, + "step": 29945 + }, + { + "epoch": 7.850589777195282, + "grad_norm": 2.003506660461426, + "learning_rate": 6.7142075271988705e-06, + "loss": 0.1939, + "num_input_tokens_seen": 8031144, + "step": 29950 + }, + { + "epoch": 7.851900393184797, + "grad_norm": 3.770963668823242, + "learning_rate": 6.706410190156945e-06, + "loss": 0.1934, + "num_input_tokens_seen": 8032536, + "step": 29955 + }, + { + "epoch": 7.853211009174312, + "grad_norm": 2.1924986839294434, + "learning_rate": 6.698616681931754e-06, + "loss": 0.2379, + "num_input_tokens_seen": 8033672, + "step": 29960 + }, + { + "epoch": 7.854521625163827, + "grad_norm": 3.008664608001709, + "learning_rate": 6.6908270041544815e-06, + "loss": 0.284, + "num_input_tokens_seen": 8035112, + "step": 29965 + }, + { + "epoch": 7.855832241153342, + "grad_norm": 2.543172597885132, + "learning_rate": 6.683041158455472e-06, + "loss": 0.2982, + "num_input_tokens_seen": 8036264, + "step": 29970 + }, + { + "epoch": 7.857142857142857, + "grad_norm": 4.751651287078857, + "learning_rate": 6.675259146464291e-06, + "loss": 0.1454, + "num_input_tokens_seen": 8037768, + "step": 29975 + }, + { + "epoch": 7.858453473132372, + "grad_norm": 2.40537166595459, + "learning_rate": 6.667480969809714e-06, + "loss": 0.2075, + "num_input_tokens_seen": 8038840, + "step": 29980 + }, + { + "epoch": 7.859764089121887, + "grad_norm": 1.4408801794052124, + "learning_rate": 6.659706630119697e-06, + "loss": 0.1442, + "num_input_tokens_seen": 8040184, + "step": 29985 + }, + { + "epoch": 7.8610747051114025, + "grad_norm": 1.4398552179336548, + "learning_rate": 6.651936129021391e-06, + "loss": 0.1478, + "num_input_tokens_seen": 8041576, + "step": 29990 + }, + { + "epoch": 7.862385321100917, + "grad_norm": 1.0309840440750122, + "learning_rate": 6.644169468141148e-06, + "loss": 0.1476, + "num_input_tokens_seen": 8043064, + "step": 29995 + }, + { + "epoch": 7.863695937090433, + "grad_norm": 3.9412896633148193, + "learning_rate": 6.636406649104529e-06, + "loss": 0.3813, + "num_input_tokens_seen": 8044536, + "step": 30000 + }, + { + "epoch": 7.865006553079947, + "grad_norm": 1.2790395021438599, + "learning_rate": 6.628647673536262e-06, + "loss": 0.1835, + "num_input_tokens_seen": 8045944, + "step": 30005 + }, + { + "epoch": 7.866317169069463, + "grad_norm": 2.723968505859375, + "learning_rate": 6.620892543060292e-06, + "loss": 0.198, + "num_input_tokens_seen": 8047096, + "step": 30010 + }, + { + "epoch": 7.8676277850589775, + "grad_norm": 0.25432947278022766, + "learning_rate": 6.613141259299754e-06, + "loss": 0.1048, + "num_input_tokens_seen": 8048344, + "step": 30015 + }, + { + "epoch": 7.868938401048493, + "grad_norm": 1.5084513425827026, + "learning_rate": 6.6053938238769845e-06, + "loss": 0.1607, + "num_input_tokens_seen": 8049320, + "step": 30020 + }, + { + "epoch": 7.870249017038008, + "grad_norm": 1.703540563583374, + "learning_rate": 6.5976502384135025e-06, + "loss": 0.191, + "num_input_tokens_seen": 8050392, + "step": 30025 + }, + { + "epoch": 7.871559633027523, + "grad_norm": 1.6506032943725586, + "learning_rate": 6.58991050453002e-06, + "loss": 0.2404, + "num_input_tokens_seen": 8052744, + "step": 30030 + }, + { + "epoch": 7.872870249017038, + "grad_norm": 2.4638466835021973, + "learning_rate": 6.582174623846477e-06, + "loss": 0.1696, + "num_input_tokens_seen": 8054056, + "step": 30035 + }, + { + "epoch": 7.8741808650065535, + "grad_norm": 1.8538662195205688, + "learning_rate": 6.57444259798195e-06, + "loss": 0.1702, + "num_input_tokens_seen": 8055208, + "step": 30040 + }, + { + "epoch": 7.875491480996068, + "grad_norm": 2.7356033325195312, + "learning_rate": 6.566714428554752e-06, + "loss": 0.1345, + "num_input_tokens_seen": 8056392, + "step": 30045 + }, + { + "epoch": 7.876802096985584, + "grad_norm": 1.4998916387557983, + "learning_rate": 6.558990117182376e-06, + "loss": 0.2267, + "num_input_tokens_seen": 8057752, + "step": 30050 + }, + { + "epoch": 7.878112712975098, + "grad_norm": 6.678801536560059, + "learning_rate": 6.551269665481502e-06, + "loss": 0.1664, + "num_input_tokens_seen": 8059176, + "step": 30055 + }, + { + "epoch": 7.879423328964613, + "grad_norm": 1.1488831043243408, + "learning_rate": 6.543553075068012e-06, + "loss": 0.0945, + "num_input_tokens_seen": 8060296, + "step": 30060 + }, + { + "epoch": 7.8807339449541285, + "grad_norm": 3.3963704109191895, + "learning_rate": 6.535840347556971e-06, + "loss": 0.2128, + "num_input_tokens_seen": 8061640, + "step": 30065 + }, + { + "epoch": 7.882044560943643, + "grad_norm": 3.9892189502716064, + "learning_rate": 6.528131484562644e-06, + "loss": 0.3545, + "num_input_tokens_seen": 8063192, + "step": 30070 + }, + { + "epoch": 7.883355176933159, + "grad_norm": 1.447434663772583, + "learning_rate": 6.520426487698478e-06, + "loss": 0.1619, + "num_input_tokens_seen": 8064472, + "step": 30075 + }, + { + "epoch": 7.884665792922673, + "grad_norm": 2.079885244369507, + "learning_rate": 6.512725358577121e-06, + "loss": 0.2712, + "num_input_tokens_seen": 8065896, + "step": 30080 + }, + { + "epoch": 7.885976408912189, + "grad_norm": 5.243135929107666, + "learning_rate": 6.505028098810406e-06, + "loss": 0.1848, + "num_input_tokens_seen": 8066920, + "step": 30085 + }, + { + "epoch": 7.8872870249017035, + "grad_norm": 2.946944236755371, + "learning_rate": 6.497334710009342e-06, + "loss": 0.1512, + "num_input_tokens_seen": 8068248, + "step": 30090 + }, + { + "epoch": 7.888597640891219, + "grad_norm": 7.704549789428711, + "learning_rate": 6.489645193784158e-06, + "loss": 0.3007, + "num_input_tokens_seen": 8069352, + "step": 30095 + }, + { + "epoch": 7.889908256880734, + "grad_norm": 3.071434259414673, + "learning_rate": 6.481959551744254e-06, + "loss": 0.3526, + "num_input_tokens_seen": 8070696, + "step": 30100 + }, + { + "epoch": 7.891218872870249, + "grad_norm": 2.0960445404052734, + "learning_rate": 6.474277785498217e-06, + "loss": 0.1723, + "num_input_tokens_seen": 8072024, + "step": 30105 + }, + { + "epoch": 7.892529488859764, + "grad_norm": 2.337650775909424, + "learning_rate": 6.46659989665383e-06, + "loss": 0.2456, + "num_input_tokens_seen": 8073384, + "step": 30110 + }, + { + "epoch": 7.8938401048492794, + "grad_norm": 2.2862019538879395, + "learning_rate": 6.458925886818062e-06, + "loss": 0.3133, + "num_input_tokens_seen": 8074888, + "step": 30115 + }, + { + "epoch": 7.895150720838794, + "grad_norm": 1.222840666770935, + "learning_rate": 6.451255757597078e-06, + "loss": 0.1991, + "num_input_tokens_seen": 8076456, + "step": 30120 + }, + { + "epoch": 7.89646133682831, + "grad_norm": 2.4409964084625244, + "learning_rate": 6.443589510596204e-06, + "loss": 0.2114, + "num_input_tokens_seen": 8077512, + "step": 30125 + }, + { + "epoch": 7.897771952817824, + "grad_norm": 4.371899604797363, + "learning_rate": 6.435927147419984e-06, + "loss": 0.094, + "num_input_tokens_seen": 8078392, + "step": 30130 + }, + { + "epoch": 7.89908256880734, + "grad_norm": 2.032994270324707, + "learning_rate": 6.4282686696721356e-06, + "loss": 0.2167, + "num_input_tokens_seen": 8080120, + "step": 30135 + }, + { + "epoch": 7.9003931847968545, + "grad_norm": 1.2554466724395752, + "learning_rate": 6.420614078955564e-06, + "loss": 0.1538, + "num_input_tokens_seen": 8081512, + "step": 30140 + }, + { + "epoch": 7.90170380078637, + "grad_norm": 1.1470561027526855, + "learning_rate": 6.412963376872358e-06, + "loss": 0.1741, + "num_input_tokens_seen": 8083320, + "step": 30145 + }, + { + "epoch": 7.903014416775885, + "grad_norm": 7.119148254394531, + "learning_rate": 6.405316565023806e-06, + "loss": 0.1796, + "num_input_tokens_seen": 8084584, + "step": 30150 + }, + { + "epoch": 7.904325032765399, + "grad_norm": 1.4344514608383179, + "learning_rate": 6.397673645010377e-06, + "loss": 0.1127, + "num_input_tokens_seen": 8085800, + "step": 30155 + }, + { + "epoch": 7.905635648754915, + "grad_norm": 2.418586492538452, + "learning_rate": 6.390034618431704e-06, + "loss": 0.196, + "num_input_tokens_seen": 8086952, + "step": 30160 + }, + { + "epoch": 7.9069462647444295, + "grad_norm": 2.0788567066192627, + "learning_rate": 6.382399486886631e-06, + "loss": 0.1889, + "num_input_tokens_seen": 8088296, + "step": 30165 + }, + { + "epoch": 7.908256880733945, + "grad_norm": 3.6403555870056152, + "learning_rate": 6.374768251973174e-06, + "loss": 0.1001, + "num_input_tokens_seen": 8089528, + "step": 30170 + }, + { + "epoch": 7.90956749672346, + "grad_norm": 1.8322107791900635, + "learning_rate": 6.367140915288542e-06, + "loss": 0.3033, + "num_input_tokens_seen": 8091192, + "step": 30175 + }, + { + "epoch": 7.910878112712975, + "grad_norm": 4.021765232086182, + "learning_rate": 6.3595174784291195e-06, + "loss": 0.2737, + "num_input_tokens_seen": 8093176, + "step": 30180 + }, + { + "epoch": 7.91218872870249, + "grad_norm": 0.5754711627960205, + "learning_rate": 6.351897942990481e-06, + "loss": 0.2397, + "num_input_tokens_seen": 8094408, + "step": 30185 + }, + { + "epoch": 7.913499344692005, + "grad_norm": 2.118314743041992, + "learning_rate": 6.344282310567384e-06, + "loss": 0.2564, + "num_input_tokens_seen": 8096088, + "step": 30190 + }, + { + "epoch": 7.91480996068152, + "grad_norm": 2.1680195331573486, + "learning_rate": 6.336670582753762e-06, + "loss": 0.2566, + "num_input_tokens_seen": 8097224, + "step": 30195 + }, + { + "epoch": 7.916120576671036, + "grad_norm": 0.8918970823287964, + "learning_rate": 6.329062761142748e-06, + "loss": 0.1294, + "num_input_tokens_seen": 8098216, + "step": 30200 + }, + { + "epoch": 7.91743119266055, + "grad_norm": 1.6993381977081299, + "learning_rate": 6.3214588473266225e-06, + "loss": 0.1478, + "num_input_tokens_seen": 8099400, + "step": 30205 + }, + { + "epoch": 7.918741808650066, + "grad_norm": 2.221506118774414, + "learning_rate": 6.313858842896895e-06, + "loss": 0.1646, + "num_input_tokens_seen": 8100312, + "step": 30210 + }, + { + "epoch": 7.9200524246395805, + "grad_norm": 3.2216267585754395, + "learning_rate": 6.306262749444219e-06, + "loss": 0.207, + "num_input_tokens_seen": 8101672, + "step": 30215 + }, + { + "epoch": 7.921363040629096, + "grad_norm": 2.6649773120880127, + "learning_rate": 6.298670568558454e-06, + "loss": 0.1617, + "num_input_tokens_seen": 8103000, + "step": 30220 + }, + { + "epoch": 7.922673656618611, + "grad_norm": 0.8414759039878845, + "learning_rate": 6.291082301828624e-06, + "loss": 0.1353, + "num_input_tokens_seen": 8104104, + "step": 30225 + }, + { + "epoch": 7.923984272608125, + "grad_norm": 1.8304400444030762, + "learning_rate": 6.283497950842942e-06, + "loss": 0.1883, + "num_input_tokens_seen": 8105432, + "step": 30230 + }, + { + "epoch": 7.925294888597641, + "grad_norm": 3.5655579566955566, + "learning_rate": 6.275917517188809e-06, + "loss": 0.2394, + "num_input_tokens_seen": 8106424, + "step": 30235 + }, + { + "epoch": 7.926605504587156, + "grad_norm": 1.3820174932479858, + "learning_rate": 6.2683410024527764e-06, + "loss": 0.1554, + "num_input_tokens_seen": 8107928, + "step": 30240 + }, + { + "epoch": 7.927916120576671, + "grad_norm": 2.979531764984131, + "learning_rate": 6.260768408220607e-06, + "loss": 0.2353, + "num_input_tokens_seen": 8109208, + "step": 30245 + }, + { + "epoch": 7.929226736566186, + "grad_norm": 2.5650956630706787, + "learning_rate": 6.253199736077231e-06, + "loss": 0.1276, + "num_input_tokens_seen": 8110472, + "step": 30250 + }, + { + "epoch": 7.930537352555701, + "grad_norm": 0.8019691109657288, + "learning_rate": 6.24563498760676e-06, + "loss": 0.2494, + "num_input_tokens_seen": 8111688, + "step": 30255 + }, + { + "epoch": 7.931847968545216, + "grad_norm": 1.1962655782699585, + "learning_rate": 6.238074164392471e-06, + "loss": 0.132, + "num_input_tokens_seen": 8112872, + "step": 30260 + }, + { + "epoch": 7.933158584534731, + "grad_norm": 1.288178563117981, + "learning_rate": 6.23051726801685e-06, + "loss": 0.1409, + "num_input_tokens_seen": 8113944, + "step": 30265 + }, + { + "epoch": 7.934469200524246, + "grad_norm": 1.7749667167663574, + "learning_rate": 6.22296430006154e-06, + "loss": 0.1205, + "num_input_tokens_seen": 8115384, + "step": 30270 + }, + { + "epoch": 7.935779816513762, + "grad_norm": 3.4886012077331543, + "learning_rate": 6.215415262107352e-06, + "loss": 0.3338, + "num_input_tokens_seen": 8118120, + "step": 30275 + }, + { + "epoch": 7.937090432503276, + "grad_norm": 3.4890851974487305, + "learning_rate": 6.207870155734291e-06, + "loss": 0.2079, + "num_input_tokens_seen": 8119176, + "step": 30280 + }, + { + "epoch": 7.938401048492792, + "grad_norm": 1.4350013732910156, + "learning_rate": 6.200328982521536e-06, + "loss": 0.1976, + "num_input_tokens_seen": 8120360, + "step": 30285 + }, + { + "epoch": 7.9397116644823065, + "grad_norm": 2.4944772720336914, + "learning_rate": 6.192791744047444e-06, + "loss": 0.223, + "num_input_tokens_seen": 8121528, + "step": 30290 + }, + { + "epoch": 7.941022280471822, + "grad_norm": 1.0631122589111328, + "learning_rate": 6.185258441889541e-06, + "loss": 0.207, + "num_input_tokens_seen": 8122856, + "step": 30295 + }, + { + "epoch": 7.942332896461337, + "grad_norm": 2.0964503288269043, + "learning_rate": 6.177729077624539e-06, + "loss": 0.3076, + "num_input_tokens_seen": 8124072, + "step": 30300 + }, + { + "epoch": 7.943643512450852, + "grad_norm": 4.042500972747803, + "learning_rate": 6.170203652828316e-06, + "loss": 0.2308, + "num_input_tokens_seen": 8125752, + "step": 30305 + }, + { + "epoch": 7.944954128440367, + "grad_norm": 1.3758225440979004, + "learning_rate": 6.162682169075934e-06, + "loss": 0.2995, + "num_input_tokens_seen": 8126904, + "step": 30310 + }, + { + "epoch": 7.946264744429882, + "grad_norm": 1.1980154514312744, + "learning_rate": 6.155164627941626e-06, + "loss": 0.224, + "num_input_tokens_seen": 8128168, + "step": 30315 + }, + { + "epoch": 7.947575360419397, + "grad_norm": 2.278718948364258, + "learning_rate": 6.147651030998799e-06, + "loss": 0.263, + "num_input_tokens_seen": 8129544, + "step": 30320 + }, + { + "epoch": 7.948885976408912, + "grad_norm": 2.103398323059082, + "learning_rate": 6.140141379820038e-06, + "loss": 0.1948, + "num_input_tokens_seen": 8130616, + "step": 30325 + }, + { + "epoch": 7.950196592398427, + "grad_norm": 3.7137246131896973, + "learning_rate": 6.132635675977099e-06, + "loss": 0.228, + "num_input_tokens_seen": 8131432, + "step": 30330 + }, + { + "epoch": 7.951507208387943, + "grad_norm": 2.503378391265869, + "learning_rate": 6.125133921040912e-06, + "loss": 0.0924, + "num_input_tokens_seen": 8132520, + "step": 30335 + }, + { + "epoch": 7.952817824377457, + "grad_norm": 1.7170305252075195, + "learning_rate": 6.1176361165815845e-06, + "loss": 0.1194, + "num_input_tokens_seen": 8133720, + "step": 30340 + }, + { + "epoch": 7.954128440366972, + "grad_norm": 1.3055301904678345, + "learning_rate": 6.1101422641683895e-06, + "loss": 0.1714, + "num_input_tokens_seen": 8135528, + "step": 30345 + }, + { + "epoch": 7.955439056356488, + "grad_norm": 1.7146378755569458, + "learning_rate": 6.102652365369779e-06, + "loss": 0.165, + "num_input_tokens_seen": 8137240, + "step": 30350 + }, + { + "epoch": 7.956749672346002, + "grad_norm": 0.859397828578949, + "learning_rate": 6.095166421753385e-06, + "loss": 0.2718, + "num_input_tokens_seen": 8138296, + "step": 30355 + }, + { + "epoch": 7.958060288335518, + "grad_norm": 1.0698018074035645, + "learning_rate": 6.087684434885985e-06, + "loss": 0.2091, + "num_input_tokens_seen": 8139816, + "step": 30360 + }, + { + "epoch": 7.9593709043250325, + "grad_norm": 3.547959327697754, + "learning_rate": 6.080206406333555e-06, + "loss": 0.2295, + "num_input_tokens_seen": 8140840, + "step": 30365 + }, + { + "epoch": 7.960681520314548, + "grad_norm": 2.112567663192749, + "learning_rate": 6.072732337661236e-06, + "loss": 0.1421, + "num_input_tokens_seen": 8141896, + "step": 30370 + }, + { + "epoch": 7.961992136304063, + "grad_norm": 2.0712928771972656, + "learning_rate": 6.065262230433325e-06, + "loss": 0.2207, + "num_input_tokens_seen": 8143528, + "step": 30375 + }, + { + "epoch": 7.963302752293578, + "grad_norm": 3.6452369689941406, + "learning_rate": 6.057796086213319e-06, + "loss": 0.2765, + "num_input_tokens_seen": 8144920, + "step": 30380 + }, + { + "epoch": 7.964613368283093, + "grad_norm": 2.0833699703216553, + "learning_rate": 6.050333906563865e-06, + "loss": 0.1783, + "num_input_tokens_seen": 8146392, + "step": 30385 + }, + { + "epoch": 7.965923984272608, + "grad_norm": 2.4956307411193848, + "learning_rate": 6.0428756930467876e-06, + "loss": 0.3089, + "num_input_tokens_seen": 8147768, + "step": 30390 + }, + { + "epoch": 7.967234600262123, + "grad_norm": 0.5999352335929871, + "learning_rate": 6.035421447223064e-06, + "loss": 0.2299, + "num_input_tokens_seen": 8148840, + "step": 30395 + }, + { + "epoch": 7.968545216251639, + "grad_norm": 2.8350670337677, + "learning_rate": 6.027971170652866e-06, + "loss": 0.1911, + "num_input_tokens_seen": 8150408, + "step": 30400 + }, + { + "epoch": 7.969855832241153, + "grad_norm": 1.935550570487976, + "learning_rate": 6.020524864895519e-06, + "loss": 0.2293, + "num_input_tokens_seen": 8151448, + "step": 30405 + }, + { + "epoch": 7.971166448230669, + "grad_norm": 2.4780452251434326, + "learning_rate": 6.013082531509523e-06, + "loss": 0.2263, + "num_input_tokens_seen": 8152936, + "step": 30410 + }, + { + "epoch": 7.972477064220183, + "grad_norm": 1.9704089164733887, + "learning_rate": 6.005644172052549e-06, + "loss": 0.1371, + "num_input_tokens_seen": 8153864, + "step": 30415 + }, + { + "epoch": 7.973787680209698, + "grad_norm": 2.124067544937134, + "learning_rate": 5.998209788081427e-06, + "loss": 0.2285, + "num_input_tokens_seen": 8155192, + "step": 30420 + }, + { + "epoch": 7.975098296199214, + "grad_norm": 4.526253700256348, + "learning_rate": 5.990779381152168e-06, + "loss": 0.2676, + "num_input_tokens_seen": 8156264, + "step": 30425 + }, + { + "epoch": 7.976408912188729, + "grad_norm": 1.865783452987671, + "learning_rate": 5.9833529528199344e-06, + "loss": 0.2229, + "num_input_tokens_seen": 8158104, + "step": 30430 + }, + { + "epoch": 7.977719528178244, + "grad_norm": 1.663765788078308, + "learning_rate": 5.9759305046390715e-06, + "loss": 0.355, + "num_input_tokens_seen": 8159464, + "step": 30435 + }, + { + "epoch": 7.9790301441677585, + "grad_norm": 0.9375343322753906, + "learning_rate": 5.968512038163085e-06, + "loss": 0.1808, + "num_input_tokens_seen": 8160520, + "step": 30440 + }, + { + "epoch": 7.980340760157274, + "grad_norm": 2.24161958694458, + "learning_rate": 5.9610975549446414e-06, + "loss": 0.1014, + "num_input_tokens_seen": 8161640, + "step": 30445 + }, + { + "epoch": 7.981651376146789, + "grad_norm": 2.4502110481262207, + "learning_rate": 5.953687056535584e-06, + "loss": 0.1782, + "num_input_tokens_seen": 8163064, + "step": 30450 + }, + { + "epoch": 7.982961992136304, + "grad_norm": 1.4802550077438354, + "learning_rate": 5.946280544486918e-06, + "loss": 0.1827, + "num_input_tokens_seen": 8164568, + "step": 30455 + }, + { + "epoch": 7.984272608125819, + "grad_norm": 1.4347444772720337, + "learning_rate": 5.938878020348809e-06, + "loss": 0.2733, + "num_input_tokens_seen": 8165896, + "step": 30460 + }, + { + "epoch": 7.985583224115334, + "grad_norm": 2.58872389793396, + "learning_rate": 5.931479485670599e-06, + "loss": 0.2831, + "num_input_tokens_seen": 8167128, + "step": 30465 + }, + { + "epoch": 7.986893840104849, + "grad_norm": 1.622969627380371, + "learning_rate": 5.924084942000787e-06, + "loss": 0.1604, + "num_input_tokens_seen": 8168696, + "step": 30470 + }, + { + "epoch": 7.988204456094365, + "grad_norm": 2.0362234115600586, + "learning_rate": 5.916694390887034e-06, + "loss": 0.2177, + "num_input_tokens_seen": 8170024, + "step": 30475 + }, + { + "epoch": 7.989515072083879, + "grad_norm": 2.1467533111572266, + "learning_rate": 5.909307833876168e-06, + "loss": 0.1319, + "num_input_tokens_seen": 8171496, + "step": 30480 + }, + { + "epoch": 7.990825688073395, + "grad_norm": 2.9651529788970947, + "learning_rate": 5.901925272514186e-06, + "loss": 0.2552, + "num_input_tokens_seen": 8172824, + "step": 30485 + }, + { + "epoch": 7.992136304062909, + "grad_norm": 0.9012473821640015, + "learning_rate": 5.8945467083462405e-06, + "loss": 0.3479, + "num_input_tokens_seen": 8174408, + "step": 30490 + }, + { + "epoch": 7.993446920052425, + "grad_norm": 1.8048958778381348, + "learning_rate": 5.887172142916663e-06, + "loss": 0.2436, + "num_input_tokens_seen": 8175688, + "step": 30495 + }, + { + "epoch": 7.99475753604194, + "grad_norm": 1.9682930707931519, + "learning_rate": 5.879801577768934e-06, + "loss": 0.2883, + "num_input_tokens_seen": 8178136, + "step": 30500 + }, + { + "epoch": 7.996068152031455, + "grad_norm": 2.579474449157715, + "learning_rate": 5.872435014445696e-06, + "loss": 0.195, + "num_input_tokens_seen": 8179800, + "step": 30505 + }, + { + "epoch": 7.99737876802097, + "grad_norm": 1.5922802686691284, + "learning_rate": 5.865072454488765e-06, + "loss": 0.7192, + "num_input_tokens_seen": 8181512, + "step": 30510 + }, + { + "epoch": 7.9986893840104845, + "grad_norm": 1.7853020429611206, + "learning_rate": 5.857713899439101e-06, + "loss": 0.142, + "num_input_tokens_seen": 8182424, + "step": 30515 + }, + { + "epoch": 8.0, + "grad_norm": 2.9409024715423584, + "learning_rate": 5.850359350836842e-06, + "loss": 0.2298, + "num_input_tokens_seen": 8183584, + "step": 30520 + }, + { + "epoch": 8.001310615989516, + "grad_norm": 2.657460927963257, + "learning_rate": 5.843008810221284e-06, + "loss": 0.3334, + "num_input_tokens_seen": 8185008, + "step": 30525 + }, + { + "epoch": 8.002096985583224, + "eval_loss": 0.7432249784469604, + "eval_runtime": 16.8666, + "eval_samples_per_second": 50.277, + "eval_steps_per_second": 25.138, + "num_input_tokens_seen": 8185712, + "step": 30528 + }, + { + "epoch": 8.00262123197903, + "grad_norm": 1.1779301166534424, + "learning_rate": 5.835662279130879e-06, + "loss": 0.3686, + "num_input_tokens_seen": 8186432, + "step": 30530 + }, + { + "epoch": 8.003931847968545, + "grad_norm": 2.928677797317505, + "learning_rate": 5.828319759103248e-06, + "loss": 0.1328, + "num_input_tokens_seen": 8187968, + "step": 30535 + }, + { + "epoch": 8.00524246395806, + "grad_norm": 2.9675352573394775, + "learning_rate": 5.8209812516751646e-06, + "loss": 0.1664, + "num_input_tokens_seen": 8189072, + "step": 30540 + }, + { + "epoch": 8.006553079947576, + "grad_norm": 4.861025333404541, + "learning_rate": 5.8136467583825685e-06, + "loss": 0.2013, + "num_input_tokens_seen": 8190256, + "step": 30545 + }, + { + "epoch": 8.00786369593709, + "grad_norm": 1.3095381259918213, + "learning_rate": 5.806316280760551e-06, + "loss": 0.1147, + "num_input_tokens_seen": 8191440, + "step": 30550 + }, + { + "epoch": 8.009174311926605, + "grad_norm": 1.3519574403762817, + "learning_rate": 5.7989898203433725e-06, + "loss": 0.1697, + "num_input_tokens_seen": 8192976, + "step": 30555 + }, + { + "epoch": 8.01048492791612, + "grad_norm": 2.48604679107666, + "learning_rate": 5.791667378664451e-06, + "loss": 0.1251, + "num_input_tokens_seen": 8194544, + "step": 30560 + }, + { + "epoch": 8.011795543905636, + "grad_norm": 1.3946071863174438, + "learning_rate": 5.784348957256352e-06, + "loss": 0.1575, + "num_input_tokens_seen": 8195728, + "step": 30565 + }, + { + "epoch": 8.01310615989515, + "grad_norm": 0.2569050192832947, + "learning_rate": 5.7770345576508176e-06, + "loss": 0.1739, + "num_input_tokens_seen": 8196992, + "step": 30570 + }, + { + "epoch": 8.014416775884666, + "grad_norm": 1.8284692764282227, + "learning_rate": 5.769724181378736e-06, + "loss": 0.1627, + "num_input_tokens_seen": 8198544, + "step": 30575 + }, + { + "epoch": 8.015727391874181, + "grad_norm": 3.063577890396118, + "learning_rate": 5.7624178299701546e-06, + "loss": 0.1297, + "num_input_tokens_seen": 8199504, + "step": 30580 + }, + { + "epoch": 8.017038007863697, + "grad_norm": 3.929248332977295, + "learning_rate": 5.755115504954284e-06, + "loss": 0.1884, + "num_input_tokens_seen": 8200864, + "step": 30585 + }, + { + "epoch": 8.01834862385321, + "grad_norm": 1.568535566329956, + "learning_rate": 5.747817207859491e-06, + "loss": 0.2877, + "num_input_tokens_seen": 8202000, + "step": 30590 + }, + { + "epoch": 8.019659239842726, + "grad_norm": 1.7801868915557861, + "learning_rate": 5.740522940213283e-06, + "loss": 0.1547, + "num_input_tokens_seen": 8203200, + "step": 30595 + }, + { + "epoch": 8.020969855832242, + "grad_norm": 1.6390119791030884, + "learning_rate": 5.733232703542346e-06, + "loss": 0.2441, + "num_input_tokens_seen": 8204480, + "step": 30600 + }, + { + "epoch": 8.022280471821757, + "grad_norm": 1.851217269897461, + "learning_rate": 5.725946499372506e-06, + "loss": 0.1449, + "num_input_tokens_seen": 8205616, + "step": 30605 + }, + { + "epoch": 8.02359108781127, + "grad_norm": 1.6518745422363281, + "learning_rate": 5.7186643292287705e-06, + "loss": 0.1883, + "num_input_tokens_seen": 8208176, + "step": 30610 + }, + { + "epoch": 8.024901703800786, + "grad_norm": 2.28741455078125, + "learning_rate": 5.711386194635274e-06, + "loss": 0.1682, + "num_input_tokens_seen": 8209168, + "step": 30615 + }, + { + "epoch": 8.026212319790302, + "grad_norm": 8.022172927856445, + "learning_rate": 5.704112097115316e-06, + "loss": 0.2169, + "num_input_tokens_seen": 8210464, + "step": 30620 + }, + { + "epoch": 8.027522935779816, + "grad_norm": 1.0428104400634766, + "learning_rate": 5.696842038191366e-06, + "loss": 0.1186, + "num_input_tokens_seen": 8211616, + "step": 30625 + }, + { + "epoch": 8.028833551769331, + "grad_norm": 1.7478609085083008, + "learning_rate": 5.689576019385015e-06, + "loss": 0.1649, + "num_input_tokens_seen": 8212832, + "step": 30630 + }, + { + "epoch": 8.030144167758847, + "grad_norm": 2.4507150650024414, + "learning_rate": 5.682314042217041e-06, + "loss": 0.2555, + "num_input_tokens_seen": 8214048, + "step": 30635 + }, + { + "epoch": 8.031454783748362, + "grad_norm": 1.2962090969085693, + "learning_rate": 5.675056108207355e-06, + "loss": 0.1063, + "num_input_tokens_seen": 8215328, + "step": 30640 + }, + { + "epoch": 8.032765399737876, + "grad_norm": 2.261882781982422, + "learning_rate": 5.667802218875037e-06, + "loss": 0.2083, + "num_input_tokens_seen": 8216720, + "step": 30645 + }, + { + "epoch": 8.034076015727392, + "grad_norm": 2.0615243911743164, + "learning_rate": 5.66055237573831e-06, + "loss": 0.2126, + "num_input_tokens_seen": 8218112, + "step": 30650 + }, + { + "epoch": 8.035386631716907, + "grad_norm": 1.5791645050048828, + "learning_rate": 5.6533065803145545e-06, + "loss": 0.102, + "num_input_tokens_seen": 8219552, + "step": 30655 + }, + { + "epoch": 8.036697247706423, + "grad_norm": 1.1752288341522217, + "learning_rate": 5.646064834120304e-06, + "loss": 0.1966, + "num_input_tokens_seen": 8221136, + "step": 30660 + }, + { + "epoch": 8.038007863695936, + "grad_norm": 1.803959608078003, + "learning_rate": 5.638827138671243e-06, + "loss": 0.213, + "num_input_tokens_seen": 8222496, + "step": 30665 + }, + { + "epoch": 8.039318479685452, + "grad_norm": 7.34580659866333, + "learning_rate": 5.631593495482209e-06, + "loss": 0.106, + "num_input_tokens_seen": 8224272, + "step": 30670 + }, + { + "epoch": 8.040629095674968, + "grad_norm": 1.027764916419983, + "learning_rate": 5.62436390606719e-06, + "loss": 0.0945, + "num_input_tokens_seen": 8225632, + "step": 30675 + }, + { + "epoch": 8.041939711664483, + "grad_norm": 2.00004243850708, + "learning_rate": 5.617138371939326e-06, + "loss": 0.1415, + "num_input_tokens_seen": 8226928, + "step": 30680 + }, + { + "epoch": 8.043250327653997, + "grad_norm": 2.2367379665374756, + "learning_rate": 5.6099168946109114e-06, + "loss": 0.1925, + "num_input_tokens_seen": 8228064, + "step": 30685 + }, + { + "epoch": 8.044560943643512, + "grad_norm": 2.8854458332061768, + "learning_rate": 5.602699475593387e-06, + "loss": 0.1315, + "num_input_tokens_seen": 8229168, + "step": 30690 + }, + { + "epoch": 8.045871559633028, + "grad_norm": 1.1459107398986816, + "learning_rate": 5.595486116397344e-06, + "loss": 0.1353, + "num_input_tokens_seen": 8230336, + "step": 30695 + }, + { + "epoch": 8.047182175622543, + "grad_norm": 2.5708487033843994, + "learning_rate": 5.588276818532531e-06, + "loss": 0.1671, + "num_input_tokens_seen": 8231408, + "step": 30700 + }, + { + "epoch": 8.048492791612057, + "grad_norm": 1.9309576749801636, + "learning_rate": 5.5810715835078434e-06, + "loss": 0.1225, + "num_input_tokens_seen": 8232784, + "step": 30705 + }, + { + "epoch": 8.049803407601573, + "grad_norm": 1.024091362953186, + "learning_rate": 5.573870412831314e-06, + "loss": 0.1597, + "num_input_tokens_seen": 8234336, + "step": 30710 + }, + { + "epoch": 8.051114023591088, + "grad_norm": 1.7622408866882324, + "learning_rate": 5.566673308010142e-06, + "loss": 0.1421, + "num_input_tokens_seen": 8235392, + "step": 30715 + }, + { + "epoch": 8.052424639580602, + "grad_norm": 1.763126015663147, + "learning_rate": 5.5594802705506626e-06, + "loss": 0.1945, + "num_input_tokens_seen": 8236608, + "step": 30720 + }, + { + "epoch": 8.053735255570118, + "grad_norm": 2.1366662979125977, + "learning_rate": 5.552291301958379e-06, + "loss": 0.2057, + "num_input_tokens_seen": 8238096, + "step": 30725 + }, + { + "epoch": 8.055045871559633, + "grad_norm": 1.8895442485809326, + "learning_rate": 5.545106403737921e-06, + "loss": 0.1444, + "num_input_tokens_seen": 8239280, + "step": 30730 + }, + { + "epoch": 8.056356487549149, + "grad_norm": 1.4220775365829468, + "learning_rate": 5.537925577393077e-06, + "loss": 0.1739, + "num_input_tokens_seen": 8240448, + "step": 30735 + }, + { + "epoch": 8.057667103538662, + "grad_norm": 2.530649185180664, + "learning_rate": 5.530748824426782e-06, + "loss": 0.1605, + "num_input_tokens_seen": 8241504, + "step": 30740 + }, + { + "epoch": 8.058977719528178, + "grad_norm": 1.336563229560852, + "learning_rate": 5.523576146341128e-06, + "loss": 0.1147, + "num_input_tokens_seen": 8242784, + "step": 30745 + }, + { + "epoch": 8.060288335517694, + "grad_norm": 1.3003658056259155, + "learning_rate": 5.5164075446373225e-06, + "loss": 0.1971, + "num_input_tokens_seen": 8243760, + "step": 30750 + }, + { + "epoch": 8.061598951507209, + "grad_norm": 1.3700520992279053, + "learning_rate": 5.509243020815755e-06, + "loss": 0.2547, + "num_input_tokens_seen": 8245888, + "step": 30755 + }, + { + "epoch": 8.062909567496723, + "grad_norm": 2.14408540725708, + "learning_rate": 5.502082576375947e-06, + "loss": 0.1598, + "num_input_tokens_seen": 8246896, + "step": 30760 + }, + { + "epoch": 8.064220183486238, + "grad_norm": 1.9487495422363281, + "learning_rate": 5.494926212816562e-06, + "loss": 0.186, + "num_input_tokens_seen": 8248704, + "step": 30765 + }, + { + "epoch": 8.065530799475754, + "grad_norm": 2.848771333694458, + "learning_rate": 5.4877739316354236e-06, + "loss": 0.2152, + "num_input_tokens_seen": 8250368, + "step": 30770 + }, + { + "epoch": 8.06684141546527, + "grad_norm": 1.2725589275360107, + "learning_rate": 5.4806257343294755e-06, + "loss": 0.1572, + "num_input_tokens_seen": 8252000, + "step": 30775 + }, + { + "epoch": 8.068152031454783, + "grad_norm": 0.8969968557357788, + "learning_rate": 5.473481622394849e-06, + "loss": 0.1138, + "num_input_tokens_seen": 8253248, + "step": 30780 + }, + { + "epoch": 8.069462647444299, + "grad_norm": 2.397468328475952, + "learning_rate": 5.466341597326774e-06, + "loss": 0.1512, + "num_input_tokens_seen": 8254400, + "step": 30785 + }, + { + "epoch": 8.070773263433814, + "grad_norm": 3.3309824466705322, + "learning_rate": 5.459205660619651e-06, + "loss": 0.209, + "num_input_tokens_seen": 8255648, + "step": 30790 + }, + { + "epoch": 8.07208387942333, + "grad_norm": 1.7662934064865112, + "learning_rate": 5.452073813767019e-06, + "loss": 0.1918, + "num_input_tokens_seen": 8256928, + "step": 30795 + }, + { + "epoch": 8.073394495412844, + "grad_norm": 1.6635900735855103, + "learning_rate": 5.444946058261563e-06, + "loss": 0.1662, + "num_input_tokens_seen": 8258368, + "step": 30800 + }, + { + "epoch": 8.07470511140236, + "grad_norm": 0.8408844470977783, + "learning_rate": 5.4378223955951074e-06, + "loss": 0.2392, + "num_input_tokens_seen": 8260864, + "step": 30805 + }, + { + "epoch": 8.076015727391875, + "grad_norm": 2.5779471397399902, + "learning_rate": 5.430702827258622e-06, + "loss": 0.1164, + "num_input_tokens_seen": 8261984, + "step": 30810 + }, + { + "epoch": 8.077326343381388, + "grad_norm": 0.6064673066139221, + "learning_rate": 5.423587354742224e-06, + "loss": 0.1014, + "num_input_tokens_seen": 8263424, + "step": 30815 + }, + { + "epoch": 8.078636959370904, + "grad_norm": 1.7804739475250244, + "learning_rate": 5.4164759795351655e-06, + "loss": 0.0985, + "num_input_tokens_seen": 8264672, + "step": 30820 + }, + { + "epoch": 8.07994757536042, + "grad_norm": 1.9015936851501465, + "learning_rate": 5.4093687031258554e-06, + "loss": 0.171, + "num_input_tokens_seen": 8266272, + "step": 30825 + }, + { + "epoch": 8.081258191349935, + "grad_norm": 2.700915813446045, + "learning_rate": 5.402265527001818e-06, + "loss": 0.1779, + "num_input_tokens_seen": 8267296, + "step": 30830 + }, + { + "epoch": 8.082568807339449, + "grad_norm": 1.9547748565673828, + "learning_rate": 5.3951664526497356e-06, + "loss": 0.248, + "num_input_tokens_seen": 8268656, + "step": 30835 + }, + { + "epoch": 8.083879423328964, + "grad_norm": 2.404815673828125, + "learning_rate": 5.388071481555448e-06, + "loss": 0.1269, + "num_input_tokens_seen": 8269904, + "step": 30840 + }, + { + "epoch": 8.08519003931848, + "grad_norm": 2.2319540977478027, + "learning_rate": 5.3809806152039125e-06, + "loss": 0.1724, + "num_input_tokens_seen": 8270992, + "step": 30845 + }, + { + "epoch": 8.086500655307995, + "grad_norm": 2.987769603729248, + "learning_rate": 5.373893855079235e-06, + "loss": 0.1834, + "num_input_tokens_seen": 8272352, + "step": 30850 + }, + { + "epoch": 8.08781127129751, + "grad_norm": 4.847469806671143, + "learning_rate": 5.366811202664662e-06, + "loss": 0.1692, + "num_input_tokens_seen": 8273424, + "step": 30855 + }, + { + "epoch": 8.089121887287025, + "grad_norm": 1.8549343347549438, + "learning_rate": 5.35973265944259e-06, + "loss": 0.1678, + "num_input_tokens_seen": 8274976, + "step": 30860 + }, + { + "epoch": 8.09043250327654, + "grad_norm": 2.6179842948913574, + "learning_rate": 5.352658226894527e-06, + "loss": 0.1897, + "num_input_tokens_seen": 8276384, + "step": 30865 + }, + { + "epoch": 8.091743119266056, + "grad_norm": 0.8500949740409851, + "learning_rate": 5.345587906501148e-06, + "loss": 0.1283, + "num_input_tokens_seen": 8277920, + "step": 30870 + }, + { + "epoch": 8.09305373525557, + "grad_norm": 1.7977588176727295, + "learning_rate": 5.338521699742263e-06, + "loss": 0.2138, + "num_input_tokens_seen": 8279488, + "step": 30875 + }, + { + "epoch": 8.094364351245085, + "grad_norm": 1.982163429260254, + "learning_rate": 5.331459608096815e-06, + "loss": 0.1986, + "num_input_tokens_seen": 8280912, + "step": 30880 + }, + { + "epoch": 8.0956749672346, + "grad_norm": 1.6802183389663696, + "learning_rate": 5.324401633042883e-06, + "loss": 0.2187, + "num_input_tokens_seen": 8282016, + "step": 30885 + }, + { + "epoch": 8.096985583224116, + "grad_norm": 2.2763304710388184, + "learning_rate": 5.317347776057685e-06, + "loss": 0.0858, + "num_input_tokens_seen": 8283120, + "step": 30890 + }, + { + "epoch": 8.09829619921363, + "grad_norm": 1.0634337663650513, + "learning_rate": 5.310298038617606e-06, + "loss": 0.1091, + "num_input_tokens_seen": 8284496, + "step": 30895 + }, + { + "epoch": 8.099606815203146, + "grad_norm": 4.706996440887451, + "learning_rate": 5.303252422198115e-06, + "loss": 0.2102, + "num_input_tokens_seen": 8285840, + "step": 30900 + }, + { + "epoch": 8.100917431192661, + "grad_norm": 0.5683854818344116, + "learning_rate": 5.296210928273862e-06, + "loss": 0.226, + "num_input_tokens_seen": 8287008, + "step": 30905 + }, + { + "epoch": 8.102228047182175, + "grad_norm": 3.680551052093506, + "learning_rate": 5.289173558318617e-06, + "loss": 0.1377, + "num_input_tokens_seen": 8288416, + "step": 30910 + }, + { + "epoch": 8.10353866317169, + "grad_norm": 1.3602787256240845, + "learning_rate": 5.282140313805284e-06, + "loss": 0.1044, + "num_input_tokens_seen": 8289808, + "step": 30915 + }, + { + "epoch": 8.104849279161206, + "grad_norm": 1.70335054397583, + "learning_rate": 5.275111196205917e-06, + "loss": 0.1313, + "num_input_tokens_seen": 8291184, + "step": 30920 + }, + { + "epoch": 8.106159895150721, + "grad_norm": 1.6278188228607178, + "learning_rate": 5.268086206991693e-06, + "loss": 0.1698, + "num_input_tokens_seen": 8292784, + "step": 30925 + }, + { + "epoch": 8.107470511140235, + "grad_norm": 3.7674994468688965, + "learning_rate": 5.2610653476329345e-06, + "loss": 0.1579, + "num_input_tokens_seen": 8294304, + "step": 30930 + }, + { + "epoch": 8.10878112712975, + "grad_norm": 2.131051540374756, + "learning_rate": 5.254048619599089e-06, + "loss": 0.1331, + "num_input_tokens_seen": 8295488, + "step": 30935 + }, + { + "epoch": 8.110091743119266, + "grad_norm": 2.4000821113586426, + "learning_rate": 5.247036024358759e-06, + "loss": 0.1401, + "num_input_tokens_seen": 8296544, + "step": 30940 + }, + { + "epoch": 8.111402359108782, + "grad_norm": 0.7358041405677795, + "learning_rate": 5.240027563379649e-06, + "loss": 0.2244, + "num_input_tokens_seen": 8299472, + "step": 30945 + }, + { + "epoch": 8.112712975098296, + "grad_norm": 2.0813419818878174, + "learning_rate": 5.233023238128623e-06, + "loss": 0.1349, + "num_input_tokens_seen": 8302656, + "step": 30950 + }, + { + "epoch": 8.114023591087811, + "grad_norm": 1.5315123796463013, + "learning_rate": 5.226023050071682e-06, + "loss": 0.1464, + "num_input_tokens_seen": 8304304, + "step": 30955 + }, + { + "epoch": 8.115334207077327, + "grad_norm": 0.09890972822904587, + "learning_rate": 5.219027000673954e-06, + "loss": 0.0919, + "num_input_tokens_seen": 8305344, + "step": 30960 + }, + { + "epoch": 8.116644823066842, + "grad_norm": 1.695235252380371, + "learning_rate": 5.212035091399694e-06, + "loss": 0.1472, + "num_input_tokens_seen": 8306560, + "step": 30965 + }, + { + "epoch": 8.117955439056356, + "grad_norm": 4.446918964385986, + "learning_rate": 5.205047323712298e-06, + "loss": 0.1379, + "num_input_tokens_seen": 8307888, + "step": 30970 + }, + { + "epoch": 8.119266055045872, + "grad_norm": 1.7317754030227661, + "learning_rate": 5.198063699074293e-06, + "loss": 0.1268, + "num_input_tokens_seen": 8309424, + "step": 30975 + }, + { + "epoch": 8.120576671035387, + "grad_norm": 1.3829482793807983, + "learning_rate": 5.191084218947351e-06, + "loss": 0.1211, + "num_input_tokens_seen": 8310752, + "step": 30980 + }, + { + "epoch": 8.1218872870249, + "grad_norm": 1.5087350606918335, + "learning_rate": 5.184108884792244e-06, + "loss": 0.0832, + "num_input_tokens_seen": 8311888, + "step": 30985 + }, + { + "epoch": 8.123197903014416, + "grad_norm": 3.429201126098633, + "learning_rate": 5.177137698068912e-06, + "loss": 0.2004, + "num_input_tokens_seen": 8312896, + "step": 30990 + }, + { + "epoch": 8.124508519003932, + "grad_norm": 2.2792797088623047, + "learning_rate": 5.170170660236406e-06, + "loss": 0.1941, + "num_input_tokens_seen": 8314144, + "step": 30995 + }, + { + "epoch": 8.125819134993447, + "grad_norm": 2.4914603233337402, + "learning_rate": 5.163207772752918e-06, + "loss": 0.1099, + "num_input_tokens_seen": 8315168, + "step": 31000 + }, + { + "epoch": 8.127129750982961, + "grad_norm": 2.199521541595459, + "learning_rate": 5.156249037075761e-06, + "loss": 0.2252, + "num_input_tokens_seen": 8316560, + "step": 31005 + }, + { + "epoch": 8.128440366972477, + "grad_norm": 2.390977621078491, + "learning_rate": 5.149294454661399e-06, + "loss": 0.157, + "num_input_tokens_seen": 8317904, + "step": 31010 + }, + { + "epoch": 8.129750982961992, + "grad_norm": 1.198110818862915, + "learning_rate": 5.142344026965418e-06, + "loss": 0.0746, + "num_input_tokens_seen": 8318960, + "step": 31015 + }, + { + "epoch": 8.131061598951508, + "grad_norm": 2.2893872261047363, + "learning_rate": 5.135397755442512e-06, + "loss": 0.1048, + "num_input_tokens_seen": 8320352, + "step": 31020 + }, + { + "epoch": 8.132372214941022, + "grad_norm": 1.9749854803085327, + "learning_rate": 5.1284556415465315e-06, + "loss": 0.2734, + "num_input_tokens_seen": 8322048, + "step": 31025 + }, + { + "epoch": 8.133682830930537, + "grad_norm": 1.7103421688079834, + "learning_rate": 5.121517686730451e-06, + "loss": 0.231, + "num_input_tokens_seen": 8323584, + "step": 31030 + }, + { + "epoch": 8.134993446920053, + "grad_norm": 3.046452045440674, + "learning_rate": 5.1145838924463725e-06, + "loss": 0.1263, + "num_input_tokens_seen": 8324944, + "step": 31035 + }, + { + "epoch": 8.136304062909568, + "grad_norm": 1.4156219959259033, + "learning_rate": 5.107654260145523e-06, + "loss": 0.1487, + "num_input_tokens_seen": 8326352, + "step": 31040 + }, + { + "epoch": 8.137614678899082, + "grad_norm": 0.5999950170516968, + "learning_rate": 5.100728791278267e-06, + "loss": 0.1545, + "num_input_tokens_seen": 8327664, + "step": 31045 + }, + { + "epoch": 8.138925294888598, + "grad_norm": 2.5988762378692627, + "learning_rate": 5.0938074872940915e-06, + "loss": 0.1261, + "num_input_tokens_seen": 8328848, + "step": 31050 + }, + { + "epoch": 8.140235910878113, + "grad_norm": 0.9969148635864258, + "learning_rate": 5.086890349641616e-06, + "loss": 0.1558, + "num_input_tokens_seen": 8329984, + "step": 31055 + }, + { + "epoch": 8.141546526867629, + "grad_norm": 1.755448579788208, + "learning_rate": 5.079977379768588e-06, + "loss": 0.1539, + "num_input_tokens_seen": 8331168, + "step": 31060 + }, + { + "epoch": 8.142857142857142, + "grad_norm": 2.5059456825256348, + "learning_rate": 5.073068579121862e-06, + "loss": 0.1736, + "num_input_tokens_seen": 8332496, + "step": 31065 + }, + { + "epoch": 8.144167758846658, + "grad_norm": 2.9093945026397705, + "learning_rate": 5.066163949147457e-06, + "loss": 0.239, + "num_input_tokens_seen": 8333616, + "step": 31070 + }, + { + "epoch": 8.145478374836173, + "grad_norm": 2.740029811859131, + "learning_rate": 5.059263491290495e-06, + "loss": 0.1425, + "num_input_tokens_seen": 8334928, + "step": 31075 + }, + { + "epoch": 8.146788990825687, + "grad_norm": 1.8090087175369263, + "learning_rate": 5.052367206995229e-06, + "loss": 0.1617, + "num_input_tokens_seen": 8336432, + "step": 31080 + }, + { + "epoch": 8.148099606815203, + "grad_norm": 1.2453705072402954, + "learning_rate": 5.0454750977050406e-06, + "loss": 0.1162, + "num_input_tokens_seen": 8337776, + "step": 31085 + }, + { + "epoch": 8.149410222804718, + "grad_norm": 1.3766528367996216, + "learning_rate": 5.038587164862435e-06, + "loss": 0.1523, + "num_input_tokens_seen": 8339264, + "step": 31090 + }, + { + "epoch": 8.150720838794234, + "grad_norm": 2.4323418140411377, + "learning_rate": 5.031703409909052e-06, + "loss": 0.07, + "num_input_tokens_seen": 8340592, + "step": 31095 + }, + { + "epoch": 8.152031454783748, + "grad_norm": 0.5718222856521606, + "learning_rate": 5.024823834285636e-06, + "loss": 0.139, + "num_input_tokens_seen": 8341712, + "step": 31100 + }, + { + "epoch": 8.153342070773263, + "grad_norm": 1.4791481494903564, + "learning_rate": 5.017948439432077e-06, + "loss": 0.1164, + "num_input_tokens_seen": 8343312, + "step": 31105 + }, + { + "epoch": 8.154652686762779, + "grad_norm": 2.062593698501587, + "learning_rate": 5.0110772267873836e-06, + "loss": 0.142, + "num_input_tokens_seen": 8344384, + "step": 31110 + }, + { + "epoch": 8.155963302752294, + "grad_norm": 1.9273483753204346, + "learning_rate": 5.004210197789688e-06, + "loss": 0.1493, + "num_input_tokens_seen": 8345536, + "step": 31115 + }, + { + "epoch": 8.157273918741808, + "grad_norm": 6.329298973083496, + "learning_rate": 4.997347353876242e-06, + "loss": 0.1932, + "num_input_tokens_seen": 8346592, + "step": 31120 + }, + { + "epoch": 8.158584534731324, + "grad_norm": 2.1057984828948975, + "learning_rate": 4.990488696483439e-06, + "loss": 0.1223, + "num_input_tokens_seen": 8347616, + "step": 31125 + }, + { + "epoch": 8.159895150720839, + "grad_norm": 2.4384994506835938, + "learning_rate": 4.983634227046785e-06, + "loss": 0.2464, + "num_input_tokens_seen": 8349168, + "step": 31130 + }, + { + "epoch": 8.161205766710355, + "grad_norm": 2.4876134395599365, + "learning_rate": 4.9767839470008935e-06, + "loss": 0.1661, + "num_input_tokens_seen": 8350720, + "step": 31135 + }, + { + "epoch": 8.162516382699868, + "grad_norm": 2.048626184463501, + "learning_rate": 4.969937857779528e-06, + "loss": 0.1688, + "num_input_tokens_seen": 8352208, + "step": 31140 + }, + { + "epoch": 8.163826998689384, + "grad_norm": 3.3338258266448975, + "learning_rate": 4.963095960815556e-06, + "loss": 0.1579, + "num_input_tokens_seen": 8353504, + "step": 31145 + }, + { + "epoch": 8.1651376146789, + "grad_norm": 1.3236327171325684, + "learning_rate": 4.95625825754098e-06, + "loss": 0.1863, + "num_input_tokens_seen": 8354560, + "step": 31150 + }, + { + "epoch": 8.166448230668415, + "grad_norm": 3.5610437393188477, + "learning_rate": 4.9494247493869165e-06, + "loss": 0.152, + "num_input_tokens_seen": 8355808, + "step": 31155 + }, + { + "epoch": 8.167758846657929, + "grad_norm": 1.421753168106079, + "learning_rate": 4.942595437783609e-06, + "loss": 0.1935, + "num_input_tokens_seen": 8357360, + "step": 31160 + }, + { + "epoch": 8.169069462647444, + "grad_norm": 1.2410132884979248, + "learning_rate": 4.9357703241604185e-06, + "loss": 0.1439, + "num_input_tokens_seen": 8358880, + "step": 31165 + }, + { + "epoch": 8.17038007863696, + "grad_norm": 1.5123876333236694, + "learning_rate": 4.928949409945832e-06, + "loss": 0.1851, + "num_input_tokens_seen": 8360496, + "step": 31170 + }, + { + "epoch": 8.171690694626474, + "grad_norm": 1.5958458185195923, + "learning_rate": 4.922132696567464e-06, + "loss": 0.2108, + "num_input_tokens_seen": 8361776, + "step": 31175 + }, + { + "epoch": 8.17300131061599, + "grad_norm": 2.923129081726074, + "learning_rate": 4.915320185452013e-06, + "loss": 0.1437, + "num_input_tokens_seen": 8362928, + "step": 31180 + }, + { + "epoch": 8.174311926605505, + "grad_norm": 0.9566220641136169, + "learning_rate": 4.9085118780253515e-06, + "loss": 0.1143, + "num_input_tokens_seen": 8364176, + "step": 31185 + }, + { + "epoch": 8.17562254259502, + "grad_norm": 1.470718264579773, + "learning_rate": 4.9017077757124385e-06, + "loss": 0.1201, + "num_input_tokens_seen": 8365376, + "step": 31190 + }, + { + "epoch": 8.176933158584534, + "grad_norm": 2.022834539413452, + "learning_rate": 4.89490787993736e-06, + "loss": 0.1594, + "num_input_tokens_seen": 8366400, + "step": 31195 + }, + { + "epoch": 8.17824377457405, + "grad_norm": 1.654137134552002, + "learning_rate": 4.888112192123323e-06, + "loss": 0.2943, + "num_input_tokens_seen": 8367792, + "step": 31200 + }, + { + "epoch": 8.179554390563565, + "grad_norm": 1.7411218881607056, + "learning_rate": 4.8813207136926555e-06, + "loss": 0.2473, + "num_input_tokens_seen": 8369232, + "step": 31205 + }, + { + "epoch": 8.18086500655308, + "grad_norm": 1.864843726158142, + "learning_rate": 4.8745334460668005e-06, + "loss": 0.1915, + "num_input_tokens_seen": 8370496, + "step": 31210 + }, + { + "epoch": 8.182175622542594, + "grad_norm": 1.9134628772735596, + "learning_rate": 4.867750390666326e-06, + "loss": 0.1185, + "num_input_tokens_seen": 8371488, + "step": 31215 + }, + { + "epoch": 8.18348623853211, + "grad_norm": 2.889944076538086, + "learning_rate": 4.860971548910903e-06, + "loss": 0.1436, + "num_input_tokens_seen": 8372752, + "step": 31220 + }, + { + "epoch": 8.184796854521625, + "grad_norm": 2.094071626663208, + "learning_rate": 4.854196922219337e-06, + "loss": 0.1404, + "num_input_tokens_seen": 8374160, + "step": 31225 + }, + { + "epoch": 8.186107470511141, + "grad_norm": 1.21337890625, + "learning_rate": 4.847426512009545e-06, + "loss": 0.0445, + "num_input_tokens_seen": 8375024, + "step": 31230 + }, + { + "epoch": 8.187418086500655, + "grad_norm": 2.43436598777771, + "learning_rate": 4.840660319698559e-06, + "loss": 0.2824, + "num_input_tokens_seen": 8376448, + "step": 31235 + }, + { + "epoch": 8.18872870249017, + "grad_norm": 2.552415132522583, + "learning_rate": 4.8338983467025394e-06, + "loss": 0.1949, + "num_input_tokens_seen": 8378016, + "step": 31240 + }, + { + "epoch": 8.190039318479686, + "grad_norm": 0.8812966346740723, + "learning_rate": 4.827140594436752e-06, + "loss": 0.1277, + "num_input_tokens_seen": 8379136, + "step": 31245 + }, + { + "epoch": 8.191349934469201, + "grad_norm": 1.8241147994995117, + "learning_rate": 4.82038706431559e-06, + "loss": 0.1841, + "num_input_tokens_seen": 8380688, + "step": 31250 + }, + { + "epoch": 8.192660550458715, + "grad_norm": 2.14605450630188, + "learning_rate": 4.81363775775254e-06, + "loss": 0.1002, + "num_input_tokens_seen": 8382016, + "step": 31255 + }, + { + "epoch": 8.19397116644823, + "grad_norm": 2.329423427581787, + "learning_rate": 4.80689267616023e-06, + "loss": 0.0979, + "num_input_tokens_seen": 8383056, + "step": 31260 + }, + { + "epoch": 8.195281782437746, + "grad_norm": 1.267533302307129, + "learning_rate": 4.800151820950391e-06, + "loss": 0.1975, + "num_input_tokens_seen": 8384336, + "step": 31265 + }, + { + "epoch": 8.19659239842726, + "grad_norm": 1.8551194667816162, + "learning_rate": 4.7934151935338706e-06, + "loss": 0.1506, + "num_input_tokens_seen": 8386160, + "step": 31270 + }, + { + "epoch": 8.197903014416775, + "grad_norm": 1.9247527122497559, + "learning_rate": 4.786682795320638e-06, + "loss": 0.163, + "num_input_tokens_seen": 8387600, + "step": 31275 + }, + { + "epoch": 8.199213630406291, + "grad_norm": 1.4893736839294434, + "learning_rate": 4.779954627719771e-06, + "loss": 0.1773, + "num_input_tokens_seen": 8389072, + "step": 31280 + }, + { + "epoch": 8.200524246395807, + "grad_norm": 1.9099963903427124, + "learning_rate": 4.773230692139463e-06, + "loss": 0.2063, + "num_input_tokens_seen": 8390336, + "step": 31285 + }, + { + "epoch": 8.20183486238532, + "grad_norm": 1.4692673683166504, + "learning_rate": 4.766510989987022e-06, + "loss": 0.0852, + "num_input_tokens_seen": 8391680, + "step": 31290 + }, + { + "epoch": 8.203145478374836, + "grad_norm": 2.768429756164551, + "learning_rate": 4.7597955226688666e-06, + "loss": 0.1937, + "num_input_tokens_seen": 8392784, + "step": 31295 + }, + { + "epoch": 8.204456094364351, + "grad_norm": 1.1038832664489746, + "learning_rate": 4.75308429159054e-06, + "loss": 0.18, + "num_input_tokens_seen": 8394288, + "step": 31300 + }, + { + "epoch": 8.205766710353867, + "grad_norm": 2.0456931591033936, + "learning_rate": 4.746377298156685e-06, + "loss": 0.1039, + "num_input_tokens_seen": 8395632, + "step": 31305 + }, + { + "epoch": 8.20707732634338, + "grad_norm": 1.6305322647094727, + "learning_rate": 4.739674543771066e-06, + "loss": 0.0788, + "num_input_tokens_seen": 8396816, + "step": 31310 + }, + { + "epoch": 8.208387942332896, + "grad_norm": 3.534648895263672, + "learning_rate": 4.732976029836561e-06, + "loss": 0.1882, + "num_input_tokens_seen": 8398048, + "step": 31315 + }, + { + "epoch": 8.209698558322412, + "grad_norm": 3.6092653274536133, + "learning_rate": 4.726281757755149e-06, + "loss": 0.1798, + "num_input_tokens_seen": 8399408, + "step": 31320 + }, + { + "epoch": 8.211009174311927, + "grad_norm": 1.9730727672576904, + "learning_rate": 4.719591728927938e-06, + "loss": 0.1739, + "num_input_tokens_seen": 8400992, + "step": 31325 + }, + { + "epoch": 8.212319790301441, + "grad_norm": 0.2663930356502533, + "learning_rate": 4.712905944755144e-06, + "loss": 0.1964, + "num_input_tokens_seen": 8402208, + "step": 31330 + }, + { + "epoch": 8.213630406290957, + "grad_norm": 0.8980347514152527, + "learning_rate": 4.706224406636072e-06, + "loss": 0.111, + "num_input_tokens_seen": 8403520, + "step": 31335 + }, + { + "epoch": 8.214941022280472, + "grad_norm": 1.3237125873565674, + "learning_rate": 4.699547115969172e-06, + "loss": 0.1615, + "num_input_tokens_seen": 8405040, + "step": 31340 + }, + { + "epoch": 8.216251638269988, + "grad_norm": 1.4147700071334839, + "learning_rate": 4.692874074151979e-06, + "loss": 0.2024, + "num_input_tokens_seen": 8406512, + "step": 31345 + }, + { + "epoch": 8.217562254259501, + "grad_norm": 2.26682710647583, + "learning_rate": 4.686205282581152e-06, + "loss": 0.171, + "num_input_tokens_seen": 8407584, + "step": 31350 + }, + { + "epoch": 8.218872870249017, + "grad_norm": 2.1628196239471436, + "learning_rate": 4.679540742652463e-06, + "loss": 0.1601, + "num_input_tokens_seen": 8408672, + "step": 31355 + }, + { + "epoch": 8.220183486238533, + "grad_norm": 0.9435498714447021, + "learning_rate": 4.672880455760786e-06, + "loss": 0.0902, + "num_input_tokens_seen": 8409712, + "step": 31360 + }, + { + "epoch": 8.221494102228046, + "grad_norm": 3.9798901081085205, + "learning_rate": 4.666224423300114e-06, + "loss": 0.2594, + "num_input_tokens_seen": 8411040, + "step": 31365 + }, + { + "epoch": 8.222804718217562, + "grad_norm": 1.218766212463379, + "learning_rate": 4.659572646663532e-06, + "loss": 0.0952, + "num_input_tokens_seen": 8412448, + "step": 31370 + }, + { + "epoch": 8.224115334207077, + "grad_norm": 4.280642032623291, + "learning_rate": 4.652925127243246e-06, + "loss": 0.156, + "num_input_tokens_seen": 8413568, + "step": 31375 + }, + { + "epoch": 8.225425950196593, + "grad_norm": 1.0217238664627075, + "learning_rate": 4.646281866430574e-06, + "loss": 0.3244, + "num_input_tokens_seen": 8414688, + "step": 31380 + }, + { + "epoch": 8.226736566186107, + "grad_norm": 1.662556767463684, + "learning_rate": 4.6396428656159424e-06, + "loss": 0.2225, + "num_input_tokens_seen": 8416160, + "step": 31385 + }, + { + "epoch": 8.228047182175622, + "grad_norm": 1.682940125465393, + "learning_rate": 4.633008126188876e-06, + "loss": 0.1971, + "num_input_tokens_seen": 8417344, + "step": 31390 + }, + { + "epoch": 8.229357798165138, + "grad_norm": 2.155972719192505, + "learning_rate": 4.626377649538019e-06, + "loss": 0.1911, + "num_input_tokens_seen": 8418528, + "step": 31395 + }, + { + "epoch": 8.230668414154653, + "grad_norm": 2.1800410747528076, + "learning_rate": 4.619751437051114e-06, + "loss": 0.1051, + "num_input_tokens_seen": 8419536, + "step": 31400 + }, + { + "epoch": 8.231979030144167, + "grad_norm": 1.0306401252746582, + "learning_rate": 4.613129490115023e-06, + "loss": 0.1302, + "num_input_tokens_seen": 8420800, + "step": 31405 + }, + { + "epoch": 8.233289646133683, + "grad_norm": 2.121847629547119, + "learning_rate": 4.6065118101157014e-06, + "loss": 0.1199, + "num_input_tokens_seen": 8422016, + "step": 31410 + }, + { + "epoch": 8.234600262123198, + "grad_norm": 1.7208747863769531, + "learning_rate": 4.599898398438221e-06, + "loss": 0.2083, + "num_input_tokens_seen": 8423552, + "step": 31415 + }, + { + "epoch": 8.235910878112714, + "grad_norm": 0.9546430706977844, + "learning_rate": 4.593289256466757e-06, + "loss": 0.1846, + "num_input_tokens_seen": 8425008, + "step": 31420 + }, + { + "epoch": 8.237221494102227, + "grad_norm": 1.0092447996139526, + "learning_rate": 4.586684385584592e-06, + "loss": 0.0933, + "num_input_tokens_seen": 8426256, + "step": 31425 + }, + { + "epoch": 8.238532110091743, + "grad_norm": 1.196913480758667, + "learning_rate": 4.580083787174114e-06, + "loss": 0.092, + "num_input_tokens_seen": 8427952, + "step": 31430 + }, + { + "epoch": 8.239842726081259, + "grad_norm": 2.6228456497192383, + "learning_rate": 4.573487462616815e-06, + "loss": 0.1745, + "num_input_tokens_seen": 8429120, + "step": 31435 + }, + { + "epoch": 8.241153342070774, + "grad_norm": 2.1814486980438232, + "learning_rate": 4.566895413293298e-06, + "loss": 0.0361, + "num_input_tokens_seen": 8430032, + "step": 31440 + }, + { + "epoch": 8.242463958060288, + "grad_norm": 1.8403512239456177, + "learning_rate": 4.560307640583264e-06, + "loss": 0.1567, + "num_input_tokens_seen": 8431024, + "step": 31445 + }, + { + "epoch": 8.243774574049803, + "grad_norm": 3.4514529705047607, + "learning_rate": 4.553724145865529e-06, + "loss": 0.1341, + "num_input_tokens_seen": 8432160, + "step": 31450 + }, + { + "epoch": 8.245085190039319, + "grad_norm": 5.485858917236328, + "learning_rate": 4.547144930517996e-06, + "loss": 0.123, + "num_input_tokens_seen": 8433328, + "step": 31455 + }, + { + "epoch": 8.246395806028833, + "grad_norm": 1.8282458782196045, + "learning_rate": 4.54056999591769e-06, + "loss": 0.2065, + "num_input_tokens_seen": 8435280, + "step": 31460 + }, + { + "epoch": 8.247706422018348, + "grad_norm": 2.1049747467041016, + "learning_rate": 4.533999343440728e-06, + "loss": 0.1545, + "num_input_tokens_seen": 8436688, + "step": 31465 + }, + { + "epoch": 8.249017038007864, + "grad_norm": 1.7685555219650269, + "learning_rate": 4.527432974462345e-06, + "loss": 0.1007, + "num_input_tokens_seen": 8437680, + "step": 31470 + }, + { + "epoch": 8.25032765399738, + "grad_norm": 2.503352403640747, + "learning_rate": 4.520870890356868e-06, + "loss": 0.2178, + "num_input_tokens_seen": 8438864, + "step": 31475 + }, + { + "epoch": 8.251638269986893, + "grad_norm": 1.7569241523742676, + "learning_rate": 4.5143130924977265e-06, + "loss": 0.1929, + "num_input_tokens_seen": 8440272, + "step": 31480 + }, + { + "epoch": 8.252948885976409, + "grad_norm": 2.408433675765991, + "learning_rate": 4.5077595822574645e-06, + "loss": 0.1966, + "num_input_tokens_seen": 8441424, + "step": 31485 + }, + { + "epoch": 8.254259501965924, + "grad_norm": 0.9503734707832336, + "learning_rate": 4.50121036100771e-06, + "loss": 0.1257, + "num_input_tokens_seen": 8442704, + "step": 31490 + }, + { + "epoch": 8.25557011795544, + "grad_norm": 3.0965490341186523, + "learning_rate": 4.494665430119208e-06, + "loss": 0.171, + "num_input_tokens_seen": 8443744, + "step": 31495 + }, + { + "epoch": 8.256880733944953, + "grad_norm": 1.7896710634231567, + "learning_rate": 4.488124790961798e-06, + "loss": 0.2475, + "num_input_tokens_seen": 8445648, + "step": 31500 + }, + { + "epoch": 8.258191349934469, + "grad_norm": 2.0297179222106934, + "learning_rate": 4.4815884449044275e-06, + "loss": 0.0989, + "num_input_tokens_seen": 8447072, + "step": 31505 + }, + { + "epoch": 8.259501965923985, + "grad_norm": 2.0406625270843506, + "learning_rate": 4.475056393315144e-06, + "loss": 0.1591, + "num_input_tokens_seen": 8448496, + "step": 31510 + }, + { + "epoch": 8.2608125819135, + "grad_norm": 1.4439725875854492, + "learning_rate": 4.468528637561095e-06, + "loss": 0.1652, + "num_input_tokens_seen": 8450336, + "step": 31515 + }, + { + "epoch": 8.262123197903014, + "grad_norm": 1.8999682664871216, + "learning_rate": 4.462005179008525e-06, + "loss": 0.2139, + "num_input_tokens_seen": 8452368, + "step": 31520 + }, + { + "epoch": 8.26343381389253, + "grad_norm": 2.4877190589904785, + "learning_rate": 4.455486019022781e-06, + "loss": 0.2327, + "num_input_tokens_seen": 8453760, + "step": 31525 + }, + { + "epoch": 8.264744429882045, + "grad_norm": 1.553764820098877, + "learning_rate": 4.448971158968318e-06, + "loss": 0.1366, + "num_input_tokens_seen": 8455024, + "step": 31530 + }, + { + "epoch": 8.26605504587156, + "grad_norm": 2.1713032722473145, + "learning_rate": 4.4424606002086814e-06, + "loss": 0.2513, + "num_input_tokens_seen": 8456640, + "step": 31535 + }, + { + "epoch": 8.267365661861074, + "grad_norm": 1.8878225088119507, + "learning_rate": 4.435954344106522e-06, + "loss": 0.1906, + "num_input_tokens_seen": 8457920, + "step": 31540 + }, + { + "epoch": 8.26867627785059, + "grad_norm": 2.1150736808776855, + "learning_rate": 4.429452392023584e-06, + "loss": 0.105, + "num_input_tokens_seen": 8459584, + "step": 31545 + }, + { + "epoch": 8.269986893840105, + "grad_norm": 1.8742355108261108, + "learning_rate": 4.4229547453207175e-06, + "loss": 0.1516, + "num_input_tokens_seen": 8460656, + "step": 31550 + }, + { + "epoch": 8.271297509829619, + "grad_norm": 1.6961677074432373, + "learning_rate": 4.416461405357869e-06, + "loss": 0.1897, + "num_input_tokens_seen": 8462064, + "step": 31555 + }, + { + "epoch": 8.272608125819135, + "grad_norm": 12.68641471862793, + "learning_rate": 4.409972373494084e-06, + "loss": 0.1938, + "num_input_tokens_seen": 8463264, + "step": 31560 + }, + { + "epoch": 8.27391874180865, + "grad_norm": 3.133979558944702, + "learning_rate": 4.403487651087509e-06, + "loss": 0.2561, + "num_input_tokens_seen": 8464560, + "step": 31565 + }, + { + "epoch": 8.275229357798166, + "grad_norm": 1.2986563444137573, + "learning_rate": 4.397007239495376e-06, + "loss": 0.1316, + "num_input_tokens_seen": 8465744, + "step": 31570 + }, + { + "epoch": 8.27653997378768, + "grad_norm": 1.7144817113876343, + "learning_rate": 4.390531140074028e-06, + "loss": 0.185, + "num_input_tokens_seen": 8467376, + "step": 31575 + }, + { + "epoch": 8.277850589777195, + "grad_norm": 2.4372501373291016, + "learning_rate": 4.384059354178893e-06, + "loss": 0.1525, + "num_input_tokens_seen": 8468432, + "step": 31580 + }, + { + "epoch": 8.27916120576671, + "grad_norm": 1.7865779399871826, + "learning_rate": 4.377591883164522e-06, + "loss": 0.1489, + "num_input_tokens_seen": 8469680, + "step": 31585 + }, + { + "epoch": 8.280471821756226, + "grad_norm": 2.807488203048706, + "learning_rate": 4.371128728384538e-06, + "loss": 0.2487, + "num_input_tokens_seen": 8471008, + "step": 31590 + }, + { + "epoch": 8.28178243774574, + "grad_norm": 1.3817720413208008, + "learning_rate": 4.364669891191667e-06, + "loss": 0.1395, + "num_input_tokens_seen": 8472400, + "step": 31595 + }, + { + "epoch": 8.283093053735255, + "grad_norm": 1.618431806564331, + "learning_rate": 4.358215372937738e-06, + "loss": 0.13, + "num_input_tokens_seen": 8474368, + "step": 31600 + }, + { + "epoch": 8.284403669724771, + "grad_norm": 2.1533493995666504, + "learning_rate": 4.35176517497366e-06, + "loss": 0.1174, + "num_input_tokens_seen": 8475424, + "step": 31605 + }, + { + "epoch": 8.285714285714286, + "grad_norm": 1.818063497543335, + "learning_rate": 4.34531929864945e-06, + "loss": 0.265, + "num_input_tokens_seen": 8477120, + "step": 31610 + }, + { + "epoch": 8.2870249017038, + "grad_norm": 2.3817007541656494, + "learning_rate": 4.3388777453142265e-06, + "loss": 0.2145, + "num_input_tokens_seen": 8478528, + "step": 31615 + }, + { + "epoch": 8.288335517693316, + "grad_norm": 1.1461108922958374, + "learning_rate": 4.332440516316188e-06, + "loss": 0.1358, + "num_input_tokens_seen": 8480032, + "step": 31620 + }, + { + "epoch": 8.289646133682831, + "grad_norm": 3.1245110034942627, + "learning_rate": 4.326007613002642e-06, + "loss": 0.163, + "num_input_tokens_seen": 8481280, + "step": 31625 + }, + { + "epoch": 8.290956749672347, + "grad_norm": 2.5301756858825684, + "learning_rate": 4.319579036719979e-06, + "loss": 0.1989, + "num_input_tokens_seen": 8483040, + "step": 31630 + }, + { + "epoch": 8.29226736566186, + "grad_norm": 1.9217748641967773, + "learning_rate": 4.313154788813684e-06, + "loss": 0.1678, + "num_input_tokens_seen": 8484016, + "step": 31635 + }, + { + "epoch": 8.293577981651376, + "grad_norm": 1.4792757034301758, + "learning_rate": 4.30673487062836e-06, + "loss": 0.1998, + "num_input_tokens_seen": 8484960, + "step": 31640 + }, + { + "epoch": 8.294888597640892, + "grad_norm": 3.4016127586364746, + "learning_rate": 4.300319283507667e-06, + "loss": 0.1946, + "num_input_tokens_seen": 8486112, + "step": 31645 + }, + { + "epoch": 8.296199213630405, + "grad_norm": 2.524888038635254, + "learning_rate": 4.29390802879438e-06, + "loss": 0.1518, + "num_input_tokens_seen": 8487072, + "step": 31650 + }, + { + "epoch": 8.297509829619921, + "grad_norm": 1.1735154390335083, + "learning_rate": 4.287501107830366e-06, + "loss": 0.2451, + "num_input_tokens_seen": 8488592, + "step": 31655 + }, + { + "epoch": 8.298820445609437, + "grad_norm": 1.5864663124084473, + "learning_rate": 4.2810985219565835e-06, + "loss": 0.1921, + "num_input_tokens_seen": 8490064, + "step": 31660 + }, + { + "epoch": 8.300131061598952, + "grad_norm": 0.8264198303222656, + "learning_rate": 4.27470027251308e-06, + "loss": 0.114, + "num_input_tokens_seen": 8491648, + "step": 31665 + }, + { + "epoch": 8.301441677588466, + "grad_norm": 0.8678906559944153, + "learning_rate": 4.268306360838998e-06, + "loss": 0.191, + "num_input_tokens_seen": 8492672, + "step": 31670 + }, + { + "epoch": 8.302752293577981, + "grad_norm": 2.0524063110351562, + "learning_rate": 4.261916788272571e-06, + "loss": 0.3168, + "num_input_tokens_seen": 8494080, + "step": 31675 + }, + { + "epoch": 8.304062909567497, + "grad_norm": 1.9223382472991943, + "learning_rate": 4.2555315561511326e-06, + "loss": 0.2365, + "num_input_tokens_seen": 8495568, + "step": 31680 + }, + { + "epoch": 8.305373525557012, + "grad_norm": 0.9122117757797241, + "learning_rate": 4.249150665811097e-06, + "loss": 0.1154, + "num_input_tokens_seen": 8496768, + "step": 31685 + }, + { + "epoch": 8.306684141546526, + "grad_norm": 2.273879289627075, + "learning_rate": 4.242774118587969e-06, + "loss": 0.1315, + "num_input_tokens_seen": 8498384, + "step": 31690 + }, + { + "epoch": 8.307994757536042, + "grad_norm": 1.7127032279968262, + "learning_rate": 4.236401915816346e-06, + "loss": 0.2773, + "num_input_tokens_seen": 8499872, + "step": 31695 + }, + { + "epoch": 8.309305373525557, + "grad_norm": 1.0480786561965942, + "learning_rate": 4.230034058829932e-06, + "loss": 0.0728, + "num_input_tokens_seen": 8500944, + "step": 31700 + }, + { + "epoch": 8.310615989515073, + "grad_norm": 1.917881965637207, + "learning_rate": 4.2236705489615e-06, + "loss": 0.1276, + "num_input_tokens_seen": 8502224, + "step": 31705 + }, + { + "epoch": 8.311926605504587, + "grad_norm": 2.4771907329559326, + "learning_rate": 4.2173113875429246e-06, + "loss": 0.1074, + "num_input_tokens_seen": 8503312, + "step": 31710 + }, + { + "epoch": 8.313237221494102, + "grad_norm": 2.2137837409973145, + "learning_rate": 4.2109565759051645e-06, + "loss": 0.1632, + "num_input_tokens_seen": 8504752, + "step": 31715 + }, + { + "epoch": 8.314547837483618, + "grad_norm": 2.077063798904419, + "learning_rate": 4.204606115378282e-06, + "loss": 0.1303, + "num_input_tokens_seen": 8506160, + "step": 31720 + }, + { + "epoch": 8.315858453473133, + "grad_norm": 3.148808240890503, + "learning_rate": 4.198260007291399e-06, + "loss": 0.3745, + "num_input_tokens_seen": 8507616, + "step": 31725 + }, + { + "epoch": 8.317169069462647, + "grad_norm": 1.0989391803741455, + "learning_rate": 4.191918252972754e-06, + "loss": 0.1114, + "num_input_tokens_seen": 8508784, + "step": 31730 + }, + { + "epoch": 8.318479685452163, + "grad_norm": 2.3646080493927, + "learning_rate": 4.185580853749666e-06, + "loss": 0.1978, + "num_input_tokens_seen": 8510080, + "step": 31735 + }, + { + "epoch": 8.319790301441678, + "grad_norm": 1.602455496788025, + "learning_rate": 4.179247810948539e-06, + "loss": 0.1466, + "num_input_tokens_seen": 8511488, + "step": 31740 + }, + { + "epoch": 8.321100917431192, + "grad_norm": 1.9520268440246582, + "learning_rate": 4.172919125894869e-06, + "loss": 0.2534, + "num_input_tokens_seen": 8512944, + "step": 31745 + }, + { + "epoch": 8.322411533420707, + "grad_norm": 2.7813668251037598, + "learning_rate": 4.166594799913237e-06, + "loss": 0.1292, + "num_input_tokens_seen": 8514128, + "step": 31750 + }, + { + "epoch": 8.323722149410223, + "grad_norm": 0.820569634437561, + "learning_rate": 4.160274834327327e-06, + "loss": 0.1211, + "num_input_tokens_seen": 8515760, + "step": 31755 + }, + { + "epoch": 8.325032765399738, + "grad_norm": 1.5970110893249512, + "learning_rate": 4.153959230459878e-06, + "loss": 0.2287, + "num_input_tokens_seen": 8517104, + "step": 31760 + }, + { + "epoch": 8.326343381389252, + "grad_norm": 1.5681041479110718, + "learning_rate": 4.147647989632744e-06, + "loss": 0.2092, + "num_input_tokens_seen": 8518256, + "step": 31765 + }, + { + "epoch": 8.327653997378768, + "grad_norm": 1.9556828737258911, + "learning_rate": 4.141341113166855e-06, + "loss": 0.275, + "num_input_tokens_seen": 8519600, + "step": 31770 + }, + { + "epoch": 8.328964613368283, + "grad_norm": 1.8340744972229004, + "learning_rate": 4.135038602382232e-06, + "loss": 0.1493, + "num_input_tokens_seen": 8520944, + "step": 31775 + }, + { + "epoch": 8.330275229357799, + "grad_norm": 2.437845468521118, + "learning_rate": 4.1287404585979764e-06, + "loss": 0.1155, + "num_input_tokens_seen": 8522160, + "step": 31780 + }, + { + "epoch": 8.331585845347313, + "grad_norm": 8.246888160705566, + "learning_rate": 4.122446683132281e-06, + "loss": 0.1982, + "num_input_tokens_seen": 8523808, + "step": 31785 + }, + { + "epoch": 8.332896461336828, + "grad_norm": 2.4080214500427246, + "learning_rate": 4.116157277302418e-06, + "loss": 0.4635, + "num_input_tokens_seen": 8525616, + "step": 31790 + }, + { + "epoch": 8.334207077326344, + "grad_norm": 1.208653450012207, + "learning_rate": 4.109872242424756e-06, + "loss": 0.1912, + "num_input_tokens_seen": 8527200, + "step": 31795 + }, + { + "epoch": 8.33551769331586, + "grad_norm": 2.220841407775879, + "learning_rate": 4.1035915798147415e-06, + "loss": 0.2076, + "num_input_tokens_seen": 8528720, + "step": 31800 + }, + { + "epoch": 8.336828309305373, + "grad_norm": 2.0706539154052734, + "learning_rate": 4.097315290786899e-06, + "loss": 0.1456, + "num_input_tokens_seen": 8529888, + "step": 31805 + }, + { + "epoch": 8.338138925294889, + "grad_norm": 3.6850900650024414, + "learning_rate": 4.091043376654843e-06, + "loss": 0.1191, + "num_input_tokens_seen": 8531360, + "step": 31810 + }, + { + "epoch": 8.339449541284404, + "grad_norm": 2.170306921005249, + "learning_rate": 4.084775838731286e-06, + "loss": 0.0847, + "num_input_tokens_seen": 8532656, + "step": 31815 + }, + { + "epoch": 8.34076015727392, + "grad_norm": 2.996669292449951, + "learning_rate": 4.078512678328009e-06, + "loss": 0.1356, + "num_input_tokens_seen": 8533712, + "step": 31820 + }, + { + "epoch": 8.342070773263433, + "grad_norm": 2.1961140632629395, + "learning_rate": 4.072253896755876e-06, + "loss": 0.114, + "num_input_tokens_seen": 8535008, + "step": 31825 + }, + { + "epoch": 8.343381389252949, + "grad_norm": 1.81759512424469, + "learning_rate": 4.065999495324843e-06, + "loss": 0.249, + "num_input_tokens_seen": 8536256, + "step": 31830 + }, + { + "epoch": 8.344692005242464, + "grad_norm": 2.3565077781677246, + "learning_rate": 4.059749475343949e-06, + "loss": 0.135, + "num_input_tokens_seen": 8537680, + "step": 31835 + }, + { + "epoch": 8.346002621231978, + "grad_norm": 2.09613299369812, + "learning_rate": 4.053503838121306e-06, + "loss": 0.1309, + "num_input_tokens_seen": 8539072, + "step": 31840 + }, + { + "epoch": 8.347313237221494, + "grad_norm": 2.061923027038574, + "learning_rate": 4.047262584964112e-06, + "loss": 0.2115, + "num_input_tokens_seen": 8540576, + "step": 31845 + }, + { + "epoch": 8.34862385321101, + "grad_norm": 2.154402017593384, + "learning_rate": 4.0410257171786595e-06, + "loss": 0.2374, + "num_input_tokens_seen": 8542176, + "step": 31850 + }, + { + "epoch": 8.349934469200525, + "grad_norm": 2.195920467376709, + "learning_rate": 4.034793236070308e-06, + "loss": 0.2409, + "num_input_tokens_seen": 8543600, + "step": 31855 + }, + { + "epoch": 8.351245085190039, + "grad_norm": 1.0898926258087158, + "learning_rate": 4.028565142943508e-06, + "loss": 0.1613, + "num_input_tokens_seen": 8545072, + "step": 31860 + }, + { + "epoch": 8.352555701179554, + "grad_norm": 2.296968460083008, + "learning_rate": 4.022341439101782e-06, + "loss": 0.1056, + "num_input_tokens_seen": 8545952, + "step": 31865 + }, + { + "epoch": 8.35386631716907, + "grad_norm": 1.4546067714691162, + "learning_rate": 4.016122125847751e-06, + "loss": 0.1513, + "num_input_tokens_seen": 8547168, + "step": 31870 + }, + { + "epoch": 8.355176933158585, + "grad_norm": 1.0638344287872314, + "learning_rate": 4.009907204483107e-06, + "loss": 0.1344, + "num_input_tokens_seen": 8548480, + "step": 31875 + }, + { + "epoch": 8.356487549148099, + "grad_norm": 1.6766676902770996, + "learning_rate": 4.003696676308613e-06, + "loss": 0.1594, + "num_input_tokens_seen": 8549904, + "step": 31880 + }, + { + "epoch": 8.357798165137615, + "grad_norm": 4.014249324798584, + "learning_rate": 3.997490542624124e-06, + "loss": 0.1011, + "num_input_tokens_seen": 8550896, + "step": 31885 + }, + { + "epoch": 8.35910878112713, + "grad_norm": 1.1994444131851196, + "learning_rate": 3.991288804728577e-06, + "loss": 0.1617, + "num_input_tokens_seen": 8552096, + "step": 31890 + }, + { + "epoch": 8.360419397116646, + "grad_norm": 2.7629241943359375, + "learning_rate": 3.985091463919982e-06, + "loss": 0.2105, + "num_input_tokens_seen": 8553680, + "step": 31895 + }, + { + "epoch": 8.36173001310616, + "grad_norm": 2.178722381591797, + "learning_rate": 3.978898521495433e-06, + "loss": 0.2486, + "num_input_tokens_seen": 8556080, + "step": 31900 + }, + { + "epoch": 8.363040629095675, + "grad_norm": 1.4719090461730957, + "learning_rate": 3.972709978751099e-06, + "loss": 0.1603, + "num_input_tokens_seen": 8557280, + "step": 31905 + }, + { + "epoch": 8.36435124508519, + "grad_norm": 0.22441366314888, + "learning_rate": 3.966525836982237e-06, + "loss": 0.1953, + "num_input_tokens_seen": 8558496, + "step": 31910 + }, + { + "epoch": 8.365661861074706, + "grad_norm": 2.329960584640503, + "learning_rate": 3.960346097483175e-06, + "loss": 0.1636, + "num_input_tokens_seen": 8560000, + "step": 31915 + }, + { + "epoch": 8.36697247706422, + "grad_norm": 2.0514070987701416, + "learning_rate": 3.954170761547324e-06, + "loss": 0.0994, + "num_input_tokens_seen": 8561120, + "step": 31920 + }, + { + "epoch": 8.368283093053735, + "grad_norm": 3.4792122840881348, + "learning_rate": 3.947999830467159e-06, + "loss": 0.1398, + "num_input_tokens_seen": 8562048, + "step": 31925 + }, + { + "epoch": 8.36959370904325, + "grad_norm": 2.1822755336761475, + "learning_rate": 3.941833305534259e-06, + "loss": 0.0873, + "num_input_tokens_seen": 8562992, + "step": 31930 + }, + { + "epoch": 8.370904325032765, + "grad_norm": 1.4748890399932861, + "learning_rate": 3.935671188039264e-06, + "loss": 0.1774, + "num_input_tokens_seen": 8564816, + "step": 31935 + }, + { + "epoch": 8.37221494102228, + "grad_norm": 1.4649194478988647, + "learning_rate": 3.9295134792718905e-06, + "loss": 0.2096, + "num_input_tokens_seen": 8566464, + "step": 31940 + }, + { + "epoch": 8.373525557011796, + "grad_norm": 1.9899547100067139, + "learning_rate": 3.923360180520941e-06, + "loss": 0.1592, + "num_input_tokens_seen": 8567744, + "step": 31945 + }, + { + "epoch": 8.374836173001311, + "grad_norm": 1.7031790018081665, + "learning_rate": 3.917211293074285e-06, + "loss": 0.3251, + "num_input_tokens_seen": 8569280, + "step": 31950 + }, + { + "epoch": 8.376146788990825, + "grad_norm": 3.46535587310791, + "learning_rate": 3.911066818218886e-06, + "loss": 0.2381, + "num_input_tokens_seen": 8570256, + "step": 31955 + }, + { + "epoch": 8.37745740498034, + "grad_norm": 0.13867707550525665, + "learning_rate": 3.9049267572407565e-06, + "loss": 0.2107, + "num_input_tokens_seen": 8572704, + "step": 31960 + }, + { + "epoch": 8.378768020969856, + "grad_norm": 1.5121746063232422, + "learning_rate": 3.898791111425007e-06, + "loss": 0.1676, + "num_input_tokens_seen": 8574016, + "step": 31965 + }, + { + "epoch": 8.380078636959372, + "grad_norm": 1.332912802696228, + "learning_rate": 3.892659882055819e-06, + "loss": 0.227, + "num_input_tokens_seen": 8575696, + "step": 31970 + }, + { + "epoch": 8.381389252948885, + "grad_norm": 0.47505179047584534, + "learning_rate": 3.886533070416445e-06, + "loss": 0.2054, + "num_input_tokens_seen": 8577248, + "step": 31975 + }, + { + "epoch": 8.382699868938401, + "grad_norm": 3.177370548248291, + "learning_rate": 3.880410677789214e-06, + "loss": 0.1642, + "num_input_tokens_seen": 8578192, + "step": 31980 + }, + { + "epoch": 8.384010484927916, + "grad_norm": 1.310563087463379, + "learning_rate": 3.87429270545554e-06, + "loss": 0.0693, + "num_input_tokens_seen": 8579488, + "step": 31985 + }, + { + "epoch": 8.385321100917432, + "grad_norm": 2.0420072078704834, + "learning_rate": 3.868179154695909e-06, + "loss": 0.1037, + "num_input_tokens_seen": 8580432, + "step": 31990 + }, + { + "epoch": 8.386631716906946, + "grad_norm": 1.5923584699630737, + "learning_rate": 3.86207002678986e-06, + "loss": 0.3002, + "num_input_tokens_seen": 8582064, + "step": 31995 + }, + { + "epoch": 8.387942332896461, + "grad_norm": 4.850085258483887, + "learning_rate": 3.855965323016031e-06, + "loss": 0.2223, + "num_input_tokens_seen": 8583424, + "step": 32000 + }, + { + "epoch": 8.389252948885977, + "grad_norm": 2.052527904510498, + "learning_rate": 3.8498650446521264e-06, + "loss": 0.197, + "num_input_tokens_seen": 8584992, + "step": 32005 + }, + { + "epoch": 8.390563564875492, + "grad_norm": 1.1550774574279785, + "learning_rate": 3.843769192974925e-06, + "loss": 0.1435, + "num_input_tokens_seen": 8586080, + "step": 32010 + }, + { + "epoch": 8.391874180865006, + "grad_norm": 1.0648080110549927, + "learning_rate": 3.837677769260273e-06, + "loss": 0.1131, + "num_input_tokens_seen": 8587424, + "step": 32015 + }, + { + "epoch": 8.393184796854522, + "grad_norm": 1.3162482976913452, + "learning_rate": 3.8315907747831e-06, + "loss": 0.081, + "num_input_tokens_seen": 8588720, + "step": 32020 + }, + { + "epoch": 8.394495412844037, + "grad_norm": 2.553898811340332, + "learning_rate": 3.825508210817402e-06, + "loss": 0.1513, + "num_input_tokens_seen": 8590144, + "step": 32025 + }, + { + "epoch": 8.395806028833551, + "grad_norm": 2.28540301322937, + "learning_rate": 3.8194300786362465e-06, + "loss": 0.1241, + "num_input_tokens_seen": 8591520, + "step": 32030 + }, + { + "epoch": 8.397116644823067, + "grad_norm": 1.598999261856079, + "learning_rate": 3.8133563795117816e-06, + "loss": 0.1879, + "num_input_tokens_seen": 8592816, + "step": 32035 + }, + { + "epoch": 8.398427260812582, + "grad_norm": 2.295132637023926, + "learning_rate": 3.807287114715216e-06, + "loss": 0.1625, + "num_input_tokens_seen": 8593824, + "step": 32040 + }, + { + "epoch": 8.399737876802098, + "grad_norm": 3.616952896118164, + "learning_rate": 3.801222285516842e-06, + "loss": 0.2191, + "num_input_tokens_seen": 8595040, + "step": 32045 + }, + { + "epoch": 8.401048492791611, + "grad_norm": 4.01534366607666, + "learning_rate": 3.7951618931860112e-06, + "loss": 0.1912, + "num_input_tokens_seen": 8596240, + "step": 32050 + }, + { + "epoch": 8.402359108781127, + "grad_norm": 2.0795650482177734, + "learning_rate": 3.789105938991161e-06, + "loss": 0.2315, + "num_input_tokens_seen": 8597424, + "step": 32055 + }, + { + "epoch": 8.403669724770642, + "grad_norm": 0.1777333915233612, + "learning_rate": 3.7830544241997864e-06, + "loss": 0.1261, + "num_input_tokens_seen": 8598512, + "step": 32060 + }, + { + "epoch": 8.404980340760158, + "grad_norm": 0.636671781539917, + "learning_rate": 3.7770073500784596e-06, + "loss": 0.1131, + "num_input_tokens_seen": 8599568, + "step": 32065 + }, + { + "epoch": 8.406290956749672, + "grad_norm": 2.0805935859680176, + "learning_rate": 3.7709647178928246e-06, + "loss": 0.1533, + "num_input_tokens_seen": 8600752, + "step": 32070 + }, + { + "epoch": 8.407601572739187, + "grad_norm": 3.223011016845703, + "learning_rate": 3.7649265289076003e-06, + "loss": 0.1156, + "num_input_tokens_seen": 8601632, + "step": 32075 + }, + { + "epoch": 8.408912188728703, + "grad_norm": 1.3480641841888428, + "learning_rate": 3.7588927843865556e-06, + "loss": 0.247, + "num_input_tokens_seen": 8603216, + "step": 32080 + }, + { + "epoch": 8.410222804718218, + "grad_norm": 2.4679744243621826, + "learning_rate": 3.7528634855925475e-06, + "loss": 0.1822, + "num_input_tokens_seen": 8604672, + "step": 32085 + }, + { + "epoch": 8.411533420707732, + "grad_norm": 3.3600170612335205, + "learning_rate": 3.7468386337875028e-06, + "loss": 0.0914, + "num_input_tokens_seen": 8605856, + "step": 32090 + }, + { + "epoch": 8.412844036697248, + "grad_norm": 2.8967957496643066, + "learning_rate": 3.7408182302324014e-06, + "loss": 0.1664, + "num_input_tokens_seen": 8606912, + "step": 32095 + }, + { + "epoch": 8.414154652686763, + "grad_norm": 0.9869368076324463, + "learning_rate": 3.734802276187319e-06, + "loss": 0.2208, + "num_input_tokens_seen": 8608640, + "step": 32100 + }, + { + "epoch": 8.415465268676279, + "grad_norm": 1.800517201423645, + "learning_rate": 3.7287907729113754e-06, + "loss": 0.1267, + "num_input_tokens_seen": 8609920, + "step": 32105 + }, + { + "epoch": 8.416775884665793, + "grad_norm": 2.220402956008911, + "learning_rate": 3.722783721662773e-06, + "loss": 0.1513, + "num_input_tokens_seen": 8611328, + "step": 32110 + }, + { + "epoch": 8.418086500655308, + "grad_norm": 0.5482990145683289, + "learning_rate": 3.7167811236987704e-06, + "loss": 0.0929, + "num_input_tokens_seen": 8612608, + "step": 32115 + }, + { + "epoch": 8.419397116644824, + "grad_norm": 1.1409895420074463, + "learning_rate": 3.710782980275704e-06, + "loss": 0.1942, + "num_input_tokens_seen": 8614016, + "step": 32120 + }, + { + "epoch": 8.420707732634337, + "grad_norm": 2.6571645736694336, + "learning_rate": 3.704789292648975e-06, + "loss": 0.2535, + "num_input_tokens_seen": 8615616, + "step": 32125 + }, + { + "epoch": 8.422018348623853, + "grad_norm": 4.97189474105835, + "learning_rate": 3.6988000620730516e-06, + "loss": 0.1013, + "num_input_tokens_seen": 8616832, + "step": 32130 + }, + { + "epoch": 8.423328964613368, + "grad_norm": 1.0515464544296265, + "learning_rate": 3.6928152898014684e-06, + "loss": 0.1871, + "num_input_tokens_seen": 8618560, + "step": 32135 + }, + { + "epoch": 8.424639580602884, + "grad_norm": 4.1705780029296875, + "learning_rate": 3.686834977086831e-06, + "loss": 0.186, + "num_input_tokens_seen": 8619776, + "step": 32140 + }, + { + "epoch": 8.425950196592398, + "grad_norm": 1.7047399282455444, + "learning_rate": 3.680859125180805e-06, + "loss": 0.1753, + "num_input_tokens_seen": 8621008, + "step": 32145 + }, + { + "epoch": 8.427260812581913, + "grad_norm": 1.5538225173950195, + "learning_rate": 3.674887735334126e-06, + "loss": 0.2404, + "num_input_tokens_seen": 8622928, + "step": 32150 + }, + { + "epoch": 8.428571428571429, + "grad_norm": 1.9569237232208252, + "learning_rate": 3.6689208087966e-06, + "loss": 0.1832, + "num_input_tokens_seen": 8624320, + "step": 32155 + }, + { + "epoch": 8.429882044560944, + "grad_norm": 2.1114635467529297, + "learning_rate": 3.6629583468170914e-06, + "loss": 0.2594, + "num_input_tokens_seen": 8625648, + "step": 32160 + }, + { + "epoch": 8.431192660550458, + "grad_norm": 2.2418336868286133, + "learning_rate": 3.65700035064353e-06, + "loss": 0.1145, + "num_input_tokens_seen": 8627152, + "step": 32165 + }, + { + "epoch": 8.432503276539974, + "grad_norm": 3.413557529449463, + "learning_rate": 3.6510468215229215e-06, + "loss": 0.1474, + "num_input_tokens_seen": 8628464, + "step": 32170 + }, + { + "epoch": 8.43381389252949, + "grad_norm": 1.3885952234268188, + "learning_rate": 3.6450977607013217e-06, + "loss": 0.0663, + "num_input_tokens_seen": 8629408, + "step": 32175 + }, + { + "epoch": 8.435124508519005, + "grad_norm": 1.2218817472457886, + "learning_rate": 3.6391531694238647e-06, + "loss": 0.1639, + "num_input_tokens_seen": 8630656, + "step": 32180 + }, + { + "epoch": 8.436435124508519, + "grad_norm": 2.240570545196533, + "learning_rate": 3.6332130489347387e-06, + "loss": 0.1493, + "num_input_tokens_seen": 8631936, + "step": 32185 + }, + { + "epoch": 8.437745740498034, + "grad_norm": 1.5376721620559692, + "learning_rate": 3.627277400477211e-06, + "loss": 0.2351, + "num_input_tokens_seen": 8633584, + "step": 32190 + }, + { + "epoch": 8.43905635648755, + "grad_norm": 0.7935342788696289, + "learning_rate": 3.6213462252935844e-06, + "loss": 0.12, + "num_input_tokens_seen": 8635584, + "step": 32195 + }, + { + "epoch": 8.440366972477065, + "grad_norm": 2.1077535152435303, + "learning_rate": 3.6154195246252575e-06, + "loss": 0.2931, + "num_input_tokens_seen": 8636912, + "step": 32200 + }, + { + "epoch": 8.441677588466579, + "grad_norm": 2.536442995071411, + "learning_rate": 3.609497299712672e-06, + "loss": 0.2187, + "num_input_tokens_seen": 8638112, + "step": 32205 + }, + { + "epoch": 8.442988204456094, + "grad_norm": 1.7944225072860718, + "learning_rate": 3.6035795517953357e-06, + "loss": 0.1559, + "num_input_tokens_seen": 8639568, + "step": 32210 + }, + { + "epoch": 8.44429882044561, + "grad_norm": 1.6957391500473022, + "learning_rate": 3.597666282111836e-06, + "loss": 0.1334, + "num_input_tokens_seen": 8640832, + "step": 32215 + }, + { + "epoch": 8.445609436435124, + "grad_norm": 2.570261001586914, + "learning_rate": 3.5917574918998024e-06, + "loss": 0.1827, + "num_input_tokens_seen": 8642192, + "step": 32220 + }, + { + "epoch": 8.44692005242464, + "grad_norm": 2.8350465297698975, + "learning_rate": 3.58585318239594e-06, + "loss": 0.1324, + "num_input_tokens_seen": 8643328, + "step": 32225 + }, + { + "epoch": 8.448230668414155, + "grad_norm": 0.872730553150177, + "learning_rate": 3.579953354836002e-06, + "loss": 0.1731, + "num_input_tokens_seen": 8644400, + "step": 32230 + }, + { + "epoch": 8.44954128440367, + "grad_norm": 1.8636517524719238, + "learning_rate": 3.5740580104548145e-06, + "loss": 0.1463, + "num_input_tokens_seen": 8645920, + "step": 32235 + }, + { + "epoch": 8.450851900393184, + "grad_norm": 3.6521291732788086, + "learning_rate": 3.568167150486265e-06, + "loss": 0.1523, + "num_input_tokens_seen": 8647152, + "step": 32240 + }, + { + "epoch": 8.4521625163827, + "grad_norm": 2.9177358150482178, + "learning_rate": 3.5622807761632985e-06, + "loss": 0.079, + "num_input_tokens_seen": 8648320, + "step": 32245 + }, + { + "epoch": 8.453473132372215, + "grad_norm": 1.3207951784133911, + "learning_rate": 3.5563988887179266e-06, + "loss": 0.0532, + "num_input_tokens_seen": 8649376, + "step": 32250 + }, + { + "epoch": 8.45478374836173, + "grad_norm": 1.7446602582931519, + "learning_rate": 3.5505214893812115e-06, + "loss": 0.1382, + "num_input_tokens_seen": 8651136, + "step": 32255 + }, + { + "epoch": 8.456094364351245, + "grad_norm": 2.1945877075195312, + "learning_rate": 3.5446485793832918e-06, + "loss": 0.1626, + "num_input_tokens_seen": 8652704, + "step": 32260 + }, + { + "epoch": 8.45740498034076, + "grad_norm": 1.5279988050460815, + "learning_rate": 3.5387801599533475e-06, + "loss": 0.218, + "num_input_tokens_seen": 8654976, + "step": 32265 + }, + { + "epoch": 8.458715596330276, + "grad_norm": 2.351870059967041, + "learning_rate": 3.5329162323196356e-06, + "loss": 0.3595, + "num_input_tokens_seen": 8656640, + "step": 32270 + }, + { + "epoch": 8.460026212319791, + "grad_norm": 3.2861642837524414, + "learning_rate": 3.527056797709466e-06, + "loss": 0.2213, + "num_input_tokens_seen": 8657968, + "step": 32275 + }, + { + "epoch": 8.461336828309305, + "grad_norm": 2.209643840789795, + "learning_rate": 3.5212018573492047e-06, + "loss": 0.2111, + "num_input_tokens_seen": 8660048, + "step": 32280 + }, + { + "epoch": 8.46264744429882, + "grad_norm": 1.747818946838379, + "learning_rate": 3.5153514124642822e-06, + "loss": 0.0938, + "num_input_tokens_seen": 8661280, + "step": 32285 + }, + { + "epoch": 8.463958060288336, + "grad_norm": 3.0856475830078125, + "learning_rate": 3.5095054642791862e-06, + "loss": 0.2362, + "num_input_tokens_seen": 8663200, + "step": 32290 + }, + { + "epoch": 8.46526867627785, + "grad_norm": 1.0672893524169922, + "learning_rate": 3.503664014017463e-06, + "loss": 0.0907, + "num_input_tokens_seen": 8664768, + "step": 32295 + }, + { + "epoch": 8.466579292267365, + "grad_norm": 2.2128186225891113, + "learning_rate": 3.4978270629017207e-06, + "loss": 0.1696, + "num_input_tokens_seen": 8665824, + "step": 32300 + }, + { + "epoch": 8.46788990825688, + "grad_norm": 6.2503767013549805, + "learning_rate": 3.491994612153618e-06, + "loss": 0.2285, + "num_input_tokens_seen": 8667200, + "step": 32305 + }, + { + "epoch": 8.469200524246396, + "grad_norm": 1.7115947008132935, + "learning_rate": 3.4861666629938896e-06, + "loss": 0.127, + "num_input_tokens_seen": 8668640, + "step": 32310 + }, + { + "epoch": 8.47051114023591, + "grad_norm": 4.193716526031494, + "learning_rate": 3.4803432166422983e-06, + "loss": 0.1615, + "num_input_tokens_seen": 8669696, + "step": 32315 + }, + { + "epoch": 8.471821756225426, + "grad_norm": 3.9435977935791016, + "learning_rate": 3.474524274317689e-06, + "loss": 0.1728, + "num_input_tokens_seen": 8670848, + "step": 32320 + }, + { + "epoch": 8.473132372214941, + "grad_norm": 3.0667717456817627, + "learning_rate": 3.468709837237949e-06, + "loss": 0.1396, + "num_input_tokens_seen": 8671968, + "step": 32325 + }, + { + "epoch": 8.474442988204457, + "grad_norm": 1.9672683477401733, + "learning_rate": 3.4628999066200427e-06, + "loss": 0.173, + "num_input_tokens_seen": 8672768, + "step": 32330 + }, + { + "epoch": 8.47575360419397, + "grad_norm": 0.9026180505752563, + "learning_rate": 3.4570944836799734e-06, + "loss": 0.1275, + "num_input_tokens_seen": 8674176, + "step": 32335 + }, + { + "epoch": 8.477064220183486, + "grad_norm": 2.082648992538452, + "learning_rate": 3.4512935696328045e-06, + "loss": 0.1883, + "num_input_tokens_seen": 8675696, + "step": 32340 + }, + { + "epoch": 8.478374836173002, + "grad_norm": 1.5321272611618042, + "learning_rate": 3.4454971656926655e-06, + "loss": 0.1476, + "num_input_tokens_seen": 8677280, + "step": 32345 + }, + { + "epoch": 8.479685452162517, + "grad_norm": 1.4663701057434082, + "learning_rate": 3.4397052730727214e-06, + "loss": 0.1065, + "num_input_tokens_seen": 8678400, + "step": 32350 + }, + { + "epoch": 8.48099606815203, + "grad_norm": 4.762807846069336, + "learning_rate": 3.4339178929852085e-06, + "loss": 0.1753, + "num_input_tokens_seen": 8679696, + "step": 32355 + }, + { + "epoch": 8.482306684141546, + "grad_norm": 2.59366774559021, + "learning_rate": 3.4281350266414158e-06, + "loss": 0.1876, + "num_input_tokens_seen": 8680720, + "step": 32360 + }, + { + "epoch": 8.483617300131062, + "grad_norm": 2.404853582382202, + "learning_rate": 3.422356675251692e-06, + "loss": 0.1912, + "num_input_tokens_seen": 8682160, + "step": 32365 + }, + { + "epoch": 8.484927916120578, + "grad_norm": 3.5279829502105713, + "learning_rate": 3.4165828400254314e-06, + "loss": 0.2006, + "num_input_tokens_seen": 8683728, + "step": 32370 + }, + { + "epoch": 8.486238532110091, + "grad_norm": 1.5265673398971558, + "learning_rate": 3.4108135221710885e-06, + "loss": 0.1157, + "num_input_tokens_seen": 8684704, + "step": 32375 + }, + { + "epoch": 8.487549148099607, + "grad_norm": 1.1407313346862793, + "learning_rate": 3.405048722896173e-06, + "loss": 0.192, + "num_input_tokens_seen": 8686288, + "step": 32380 + }, + { + "epoch": 8.488859764089122, + "grad_norm": 1.635536789894104, + "learning_rate": 3.3992884434072474e-06, + "loss": 0.1706, + "num_input_tokens_seen": 8687408, + "step": 32385 + }, + { + "epoch": 8.490170380078636, + "grad_norm": 11.519933700561523, + "learning_rate": 3.3935326849099253e-06, + "loss": 0.144, + "num_input_tokens_seen": 8688592, + "step": 32390 + }, + { + "epoch": 8.491480996068152, + "grad_norm": 3.0456643104553223, + "learning_rate": 3.3877814486088816e-06, + "loss": 0.1593, + "num_input_tokens_seen": 8689792, + "step": 32395 + }, + { + "epoch": 8.492791612057667, + "grad_norm": 3.517651081085205, + "learning_rate": 3.382034735707834e-06, + "loss": 0.167, + "num_input_tokens_seen": 8690896, + "step": 32400 + }, + { + "epoch": 8.494102228047183, + "grad_norm": 1.54917311668396, + "learning_rate": 3.376292547409568e-06, + "loss": 0.1035, + "num_input_tokens_seen": 8691968, + "step": 32405 + }, + { + "epoch": 8.495412844036696, + "grad_norm": 1.0798742771148682, + "learning_rate": 3.3705548849159054e-06, + "loss": 0.208, + "num_input_tokens_seen": 8693648, + "step": 32410 + }, + { + "epoch": 8.496723460026212, + "grad_norm": 1.782442331314087, + "learning_rate": 3.364821749427732e-06, + "loss": 0.1951, + "num_input_tokens_seen": 8695056, + "step": 32415 + }, + { + "epoch": 8.498034076015728, + "grad_norm": 0.779071569442749, + "learning_rate": 3.3590931421449867e-06, + "loss": 0.0694, + "num_input_tokens_seen": 8696368, + "step": 32420 + }, + { + "epoch": 8.499344692005243, + "grad_norm": 2.103813648223877, + "learning_rate": 3.3533690642666595e-06, + "loss": 0.0639, + "num_input_tokens_seen": 8697472, + "step": 32425 + }, + { + "epoch": 8.500655307994757, + "grad_norm": 2.75056791305542, + "learning_rate": 3.347649516990778e-06, + "loss": 0.1777, + "num_input_tokens_seen": 8698880, + "step": 32430 + }, + { + "epoch": 8.501965923984272, + "grad_norm": 1.7580151557922363, + "learning_rate": 3.3419345015144406e-06, + "loss": 0.1042, + "num_input_tokens_seen": 8700096, + "step": 32435 + }, + { + "epoch": 8.502228047182175, + "eval_loss": 0.8241025805473328, + "eval_runtime": 16.8629, + "eval_samples_per_second": 50.288, + "eval_steps_per_second": 25.144, + "num_input_tokens_seen": 8700416, + "step": 32436 + }, + { + "epoch": 8.503276539973788, + "grad_norm": 3.206566095352173, + "learning_rate": 3.3362240190337817e-06, + "loss": 0.3265, + "num_input_tokens_seen": 8701568, + "step": 32440 + }, + { + "epoch": 8.504587155963304, + "grad_norm": 2.0994298458099365, + "learning_rate": 3.3305180707440124e-06, + "loss": 0.1729, + "num_input_tokens_seen": 8702976, + "step": 32445 + }, + { + "epoch": 8.505897771952817, + "grad_norm": 4.565936088562012, + "learning_rate": 3.3248166578393665e-06, + "loss": 0.2445, + "num_input_tokens_seen": 8704208, + "step": 32450 + }, + { + "epoch": 8.507208387942333, + "grad_norm": 2.3401005268096924, + "learning_rate": 3.319119781513144e-06, + "loss": 0.3193, + "num_input_tokens_seen": 8705440, + "step": 32455 + }, + { + "epoch": 8.508519003931848, + "grad_norm": 3.0628364086151123, + "learning_rate": 3.313427442957695e-06, + "loss": 0.0513, + "num_input_tokens_seen": 8706368, + "step": 32460 + }, + { + "epoch": 8.509829619921362, + "grad_norm": 3.0912275314331055, + "learning_rate": 3.3077396433644046e-06, + "loss": 0.1621, + "num_input_tokens_seen": 8708048, + "step": 32465 + }, + { + "epoch": 8.511140235910878, + "grad_norm": 1.8195706605911255, + "learning_rate": 3.3020563839237267e-06, + "loss": 0.1733, + "num_input_tokens_seen": 8709200, + "step": 32470 + }, + { + "epoch": 8.512450851900393, + "grad_norm": 2.020174980163574, + "learning_rate": 3.2963776658251537e-06, + "loss": 0.1547, + "num_input_tokens_seen": 8710304, + "step": 32475 + }, + { + "epoch": 8.513761467889909, + "grad_norm": 2.7012457847595215, + "learning_rate": 3.2907034902572382e-06, + "loss": 0.3586, + "num_input_tokens_seen": 8711632, + "step": 32480 + }, + { + "epoch": 8.515072083879424, + "grad_norm": 0.3714604377746582, + "learning_rate": 3.2850338584075686e-06, + "loss": 0.0875, + "num_input_tokens_seen": 8712464, + "step": 32485 + }, + { + "epoch": 8.516382699868938, + "grad_norm": 4.464524269104004, + "learning_rate": 3.279368771462793e-06, + "loss": 0.2251, + "num_input_tokens_seen": 8713568, + "step": 32490 + }, + { + "epoch": 8.517693315858454, + "grad_norm": 4.425812721252441, + "learning_rate": 3.273708230608602e-06, + "loss": 0.2073, + "num_input_tokens_seen": 8715136, + "step": 32495 + }, + { + "epoch": 8.51900393184797, + "grad_norm": 1.5469448566436768, + "learning_rate": 3.2680522370297397e-06, + "loss": 0.1126, + "num_input_tokens_seen": 8717136, + "step": 32500 + }, + { + "epoch": 8.520314547837483, + "grad_norm": 1.8311035633087158, + "learning_rate": 3.262400791909992e-06, + "loss": 0.1323, + "num_input_tokens_seen": 8718544, + "step": 32505 + }, + { + "epoch": 8.521625163826998, + "grad_norm": 1.1365230083465576, + "learning_rate": 3.256753896432202e-06, + "loss": 0.1449, + "num_input_tokens_seen": 8719680, + "step": 32510 + }, + { + "epoch": 8.522935779816514, + "grad_norm": 2.142810106277466, + "learning_rate": 3.2511115517782494e-06, + "loss": 0.1116, + "num_input_tokens_seen": 8721344, + "step": 32515 + }, + { + "epoch": 8.52424639580603, + "grad_norm": 0.8511820435523987, + "learning_rate": 3.2454737591290695e-06, + "loss": 0.2223, + "num_input_tokens_seen": 8722752, + "step": 32520 + }, + { + "epoch": 8.525557011795543, + "grad_norm": 1.27901029586792, + "learning_rate": 3.2398405196646443e-06, + "loss": 0.0542, + "num_input_tokens_seen": 8723984, + "step": 32525 + }, + { + "epoch": 8.526867627785059, + "grad_norm": 2.491163969039917, + "learning_rate": 3.2342118345639993e-06, + "loss": 0.2308, + "num_input_tokens_seen": 8725184, + "step": 32530 + }, + { + "epoch": 8.528178243774574, + "grad_norm": 3.1925530433654785, + "learning_rate": 3.2285877050052093e-06, + "loss": 0.1459, + "num_input_tokens_seen": 8726592, + "step": 32535 + }, + { + "epoch": 8.52948885976409, + "grad_norm": 1.0883021354675293, + "learning_rate": 3.2229681321653966e-06, + "loss": 0.0786, + "num_input_tokens_seen": 8727664, + "step": 32540 + }, + { + "epoch": 8.530799475753604, + "grad_norm": 2.8069310188293457, + "learning_rate": 3.2173531172207283e-06, + "loss": 0.1868, + "num_input_tokens_seen": 8728928, + "step": 32545 + }, + { + "epoch": 8.53211009174312, + "grad_norm": 3.183591604232788, + "learning_rate": 3.211742661346409e-06, + "loss": 0.1454, + "num_input_tokens_seen": 8730256, + "step": 32550 + }, + { + "epoch": 8.533420707732635, + "grad_norm": 8.72448444366455, + "learning_rate": 3.2061367657167025e-06, + "loss": 0.1322, + "num_input_tokens_seen": 8731248, + "step": 32555 + }, + { + "epoch": 8.534731323722148, + "grad_norm": 1.5676411390304565, + "learning_rate": 3.200535431504914e-06, + "loss": 0.1815, + "num_input_tokens_seen": 8732624, + "step": 32560 + }, + { + "epoch": 8.536041939711664, + "grad_norm": 2.7178075313568115, + "learning_rate": 3.194938659883398e-06, + "loss": 0.1275, + "num_input_tokens_seen": 8733728, + "step": 32565 + }, + { + "epoch": 8.53735255570118, + "grad_norm": 2.3494982719421387, + "learning_rate": 3.189346452023542e-06, + "loss": 0.2082, + "num_input_tokens_seen": 8734784, + "step": 32570 + }, + { + "epoch": 8.538663171690695, + "grad_norm": 1.9608404636383057, + "learning_rate": 3.1837588090957853e-06, + "loss": 0.1223, + "num_input_tokens_seen": 8736320, + "step": 32575 + }, + { + "epoch": 8.539973787680209, + "grad_norm": 2.706317901611328, + "learning_rate": 3.1781757322696224e-06, + "loss": 0.1842, + "num_input_tokens_seen": 8737424, + "step": 32580 + }, + { + "epoch": 8.541284403669724, + "grad_norm": 2.6715033054351807, + "learning_rate": 3.172597222713569e-06, + "loss": 0.1449, + "num_input_tokens_seen": 8738592, + "step": 32585 + }, + { + "epoch": 8.54259501965924, + "grad_norm": 5.568932056427002, + "learning_rate": 3.1670232815951994e-06, + "loss": 0.1205, + "num_input_tokens_seen": 8739936, + "step": 32590 + }, + { + "epoch": 8.543905635648755, + "grad_norm": 2.7953367233276367, + "learning_rate": 3.1614539100811356e-06, + "loss": 0.2535, + "num_input_tokens_seen": 8741440, + "step": 32595 + }, + { + "epoch": 8.54521625163827, + "grad_norm": 1.6595025062561035, + "learning_rate": 3.1558891093370318e-06, + "loss": 0.2241, + "num_input_tokens_seen": 8743024, + "step": 32600 + }, + { + "epoch": 8.546526867627785, + "grad_norm": 0.08479997515678406, + "learning_rate": 3.150328880527595e-06, + "loss": 0.128, + "num_input_tokens_seen": 8744160, + "step": 32605 + }, + { + "epoch": 8.5478374836173, + "grad_norm": 1.4804718494415283, + "learning_rate": 3.1447732248165633e-06, + "loss": 0.137, + "num_input_tokens_seen": 8745344, + "step": 32610 + }, + { + "epoch": 8.549148099606816, + "grad_norm": 1.4827264547348022, + "learning_rate": 3.139222143366746e-06, + "loss": 0.1581, + "num_input_tokens_seen": 8746800, + "step": 32615 + }, + { + "epoch": 8.55045871559633, + "grad_norm": 1.568688154220581, + "learning_rate": 3.1336756373399556e-06, + "loss": 0.1209, + "num_input_tokens_seen": 8748416, + "step": 32620 + }, + { + "epoch": 8.551769331585845, + "grad_norm": 1.704918622970581, + "learning_rate": 3.128133707897071e-06, + "loss": 0.1772, + "num_input_tokens_seen": 8749808, + "step": 32625 + }, + { + "epoch": 8.55307994757536, + "grad_norm": 0.1786339432001114, + "learning_rate": 3.12259635619801e-06, + "loss": 0.2522, + "num_input_tokens_seen": 8751008, + "step": 32630 + }, + { + "epoch": 8.554390563564876, + "grad_norm": 1.1163657903671265, + "learning_rate": 3.1170635834017303e-06, + "loss": 0.1745, + "num_input_tokens_seen": 8752416, + "step": 32635 + }, + { + "epoch": 8.55570117955439, + "grad_norm": 2.0079615116119385, + "learning_rate": 3.1115353906662313e-06, + "loss": 0.1155, + "num_input_tokens_seen": 8753456, + "step": 32640 + }, + { + "epoch": 8.557011795543906, + "grad_norm": 2.798100709915161, + "learning_rate": 3.1060117791485544e-06, + "loss": 0.1362, + "num_input_tokens_seen": 8754768, + "step": 32645 + }, + { + "epoch": 8.558322411533421, + "grad_norm": 1.5590604543685913, + "learning_rate": 3.1004927500047827e-06, + "loss": 0.1794, + "num_input_tokens_seen": 8756160, + "step": 32650 + }, + { + "epoch": 8.559633027522935, + "grad_norm": 3.3241946697235107, + "learning_rate": 3.0949783043900393e-06, + "loss": 0.1211, + "num_input_tokens_seen": 8757088, + "step": 32655 + }, + { + "epoch": 8.56094364351245, + "grad_norm": 1.7355921268463135, + "learning_rate": 3.0894684434584958e-06, + "loss": 0.2267, + "num_input_tokens_seen": 8758464, + "step": 32660 + }, + { + "epoch": 8.562254259501966, + "grad_norm": 2.7072670459747314, + "learning_rate": 3.083963168363341e-06, + "loss": 0.1597, + "num_input_tokens_seen": 8759792, + "step": 32665 + }, + { + "epoch": 8.563564875491481, + "grad_norm": 1.5645323991775513, + "learning_rate": 3.0784624802568192e-06, + "loss": 0.1962, + "num_input_tokens_seen": 8761248, + "step": 32670 + }, + { + "epoch": 8.564875491480995, + "grad_norm": 3.8627238273620605, + "learning_rate": 3.072966380290232e-06, + "loss": 0.1562, + "num_input_tokens_seen": 8762480, + "step": 32675 + }, + { + "epoch": 8.56618610747051, + "grad_norm": 2.0908026695251465, + "learning_rate": 3.0674748696138927e-06, + "loss": 0.2542, + "num_input_tokens_seen": 8764176, + "step": 32680 + }, + { + "epoch": 8.567496723460026, + "grad_norm": 1.2099167108535767, + "learning_rate": 3.0619879493771654e-06, + "loss": 0.2891, + "num_input_tokens_seen": 8765344, + "step": 32685 + }, + { + "epoch": 8.568807339449542, + "grad_norm": 1.3263916969299316, + "learning_rate": 3.056505620728456e-06, + "loss": 0.0718, + "num_input_tokens_seen": 8766608, + "step": 32690 + }, + { + "epoch": 8.570117955439056, + "grad_norm": 1.5009233951568604, + "learning_rate": 3.051027884815211e-06, + "loss": 0.2521, + "num_input_tokens_seen": 8768576, + "step": 32695 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 2.7780864238739014, + "learning_rate": 3.0455547427838986e-06, + "loss": 0.1184, + "num_input_tokens_seen": 8769712, + "step": 32700 + }, + { + "epoch": 8.572739187418087, + "grad_norm": 1.1477315425872803, + "learning_rate": 3.0400861957800423e-06, + "loss": 0.2782, + "num_input_tokens_seen": 8771792, + "step": 32705 + }, + { + "epoch": 8.574049803407602, + "grad_norm": 1.4016242027282715, + "learning_rate": 3.034622244948204e-06, + "loss": 0.1918, + "num_input_tokens_seen": 8772912, + "step": 32710 + }, + { + "epoch": 8.575360419397116, + "grad_norm": 0.9197458028793335, + "learning_rate": 3.029162891431975e-06, + "loss": 0.1364, + "num_input_tokens_seen": 8774656, + "step": 32715 + }, + { + "epoch": 8.576671035386632, + "grad_norm": 1.6606436967849731, + "learning_rate": 3.02370813637399e-06, + "loss": 0.1207, + "num_input_tokens_seen": 8776688, + "step": 32720 + }, + { + "epoch": 8.577981651376147, + "grad_norm": 1.679185390472412, + "learning_rate": 3.018257980915917e-06, + "loss": 0.0862, + "num_input_tokens_seen": 8777680, + "step": 32725 + }, + { + "epoch": 8.579292267365663, + "grad_norm": 2.896648406982422, + "learning_rate": 3.0128124261984693e-06, + "loss": 0.2107, + "num_input_tokens_seen": 8778656, + "step": 32730 + }, + { + "epoch": 8.580602883355176, + "grad_norm": 0.8088776469230652, + "learning_rate": 3.007371473361398e-06, + "loss": 0.1003, + "num_input_tokens_seen": 8779968, + "step": 32735 + }, + { + "epoch": 8.581913499344692, + "grad_norm": 1.459302306175232, + "learning_rate": 3.00193512354347e-06, + "loss": 0.2009, + "num_input_tokens_seen": 8781888, + "step": 32740 + }, + { + "epoch": 8.583224115334207, + "grad_norm": 4.5827155113220215, + "learning_rate": 2.996503377882512e-06, + "loss": 0.1956, + "num_input_tokens_seen": 8783424, + "step": 32745 + }, + { + "epoch": 8.584534731323721, + "grad_norm": 7.742435932159424, + "learning_rate": 2.9910762375153752e-06, + "loss": 0.0765, + "num_input_tokens_seen": 8784496, + "step": 32750 + }, + { + "epoch": 8.585845347313237, + "grad_norm": 3.788538932800293, + "learning_rate": 2.9856537035779564e-06, + "loss": 0.2852, + "num_input_tokens_seen": 8785840, + "step": 32755 + }, + { + "epoch": 8.587155963302752, + "grad_norm": 2.40543532371521, + "learning_rate": 2.980235777205179e-06, + "loss": 0.126, + "num_input_tokens_seen": 8787392, + "step": 32760 + }, + { + "epoch": 8.588466579292268, + "grad_norm": 2.150217056274414, + "learning_rate": 2.9748224595310042e-06, + "loss": 0.1326, + "num_input_tokens_seen": 8788704, + "step": 32765 + }, + { + "epoch": 8.589777195281782, + "grad_norm": 2.5120692253112793, + "learning_rate": 2.9694137516884353e-06, + "loss": 0.1418, + "num_input_tokens_seen": 8789984, + "step": 32770 + }, + { + "epoch": 8.591087811271297, + "grad_norm": 3.545355796813965, + "learning_rate": 2.9640096548094992e-06, + "loss": 0.1859, + "num_input_tokens_seen": 8791792, + "step": 32775 + }, + { + "epoch": 8.592398427260813, + "grad_norm": 0.7918424010276794, + "learning_rate": 2.958610170025275e-06, + "loss": 0.1592, + "num_input_tokens_seen": 8793152, + "step": 32780 + }, + { + "epoch": 8.593709043250328, + "grad_norm": 2.062091588973999, + "learning_rate": 2.9532152984658446e-06, + "loss": 0.1679, + "num_input_tokens_seen": 8794432, + "step": 32785 + }, + { + "epoch": 8.595019659239842, + "grad_norm": 0.6146033406257629, + "learning_rate": 2.9478250412603657e-06, + "loss": 0.0785, + "num_input_tokens_seen": 8795648, + "step": 32790 + }, + { + "epoch": 8.596330275229358, + "grad_norm": 0.8435283899307251, + "learning_rate": 2.942439399537e-06, + "loss": 0.0916, + "num_input_tokens_seen": 8796640, + "step": 32795 + }, + { + "epoch": 8.597640891218873, + "grad_norm": 3.6023919582366943, + "learning_rate": 2.9370583744229572e-06, + "loss": 0.3351, + "num_input_tokens_seen": 8797744, + "step": 32800 + }, + { + "epoch": 8.598951507208389, + "grad_norm": 2.166363000869751, + "learning_rate": 2.931681967044475e-06, + "loss": 0.1012, + "num_input_tokens_seen": 8798800, + "step": 32805 + }, + { + "epoch": 8.600262123197902, + "grad_norm": 2.972397565841675, + "learning_rate": 2.9263101785268254e-06, + "loss": 0.1483, + "num_input_tokens_seen": 8800016, + "step": 32810 + }, + { + "epoch": 8.601572739187418, + "grad_norm": 3.6610357761383057, + "learning_rate": 2.9209430099943204e-06, + "loss": 0.1489, + "num_input_tokens_seen": 8801488, + "step": 32815 + }, + { + "epoch": 8.602883355176933, + "grad_norm": 2.7744994163513184, + "learning_rate": 2.915580462570289e-06, + "loss": 0.143, + "num_input_tokens_seen": 8802592, + "step": 32820 + }, + { + "epoch": 8.604193971166449, + "grad_norm": 3.969348907470703, + "learning_rate": 2.910222537377108e-06, + "loss": 0.2114, + "num_input_tokens_seen": 8803872, + "step": 32825 + }, + { + "epoch": 8.605504587155963, + "grad_norm": 1.3093658685684204, + "learning_rate": 2.904869235536181e-06, + "loss": 0.1366, + "num_input_tokens_seen": 8805008, + "step": 32830 + }, + { + "epoch": 8.606815203145478, + "grad_norm": 8.833200454711914, + "learning_rate": 2.8995205581679452e-06, + "loss": 0.0634, + "num_input_tokens_seen": 8806000, + "step": 32835 + }, + { + "epoch": 8.608125819134994, + "grad_norm": 4.469554901123047, + "learning_rate": 2.8941765063918657e-06, + "loss": 0.131, + "num_input_tokens_seen": 8807248, + "step": 32840 + }, + { + "epoch": 8.609436435124508, + "grad_norm": 1.0593615770339966, + "learning_rate": 2.888837081326451e-06, + "loss": 0.1142, + "num_input_tokens_seen": 8808752, + "step": 32845 + }, + { + "epoch": 8.610747051114023, + "grad_norm": 2.4046406745910645, + "learning_rate": 2.8835022840892373e-06, + "loss": 0.093, + "num_input_tokens_seen": 8810096, + "step": 32850 + }, + { + "epoch": 8.612057667103539, + "grad_norm": 1.8283796310424805, + "learning_rate": 2.8781721157967734e-06, + "loss": 0.3095, + "num_input_tokens_seen": 8812144, + "step": 32855 + }, + { + "epoch": 8.613368283093054, + "grad_norm": 2.874671697616577, + "learning_rate": 2.872846577564664e-06, + "loss": 0.2331, + "num_input_tokens_seen": 8813200, + "step": 32860 + }, + { + "epoch": 8.614678899082568, + "grad_norm": 1.703589677810669, + "learning_rate": 2.8675256705075343e-06, + "loss": 0.1637, + "num_input_tokens_seen": 8814496, + "step": 32865 + }, + { + "epoch": 8.615989515072084, + "grad_norm": 4.139475345611572, + "learning_rate": 2.8622093957390377e-06, + "loss": 0.2116, + "num_input_tokens_seen": 8815600, + "step": 32870 + }, + { + "epoch": 8.617300131061599, + "grad_norm": 3.248162269592285, + "learning_rate": 2.856897754371865e-06, + "loss": 0.2577, + "num_input_tokens_seen": 8816672, + "step": 32875 + }, + { + "epoch": 8.618610747051115, + "grad_norm": 2.1044280529022217, + "learning_rate": 2.8515907475177293e-06, + "loss": 0.1181, + "num_input_tokens_seen": 8817824, + "step": 32880 + }, + { + "epoch": 8.619921363040628, + "grad_norm": 2.512974500656128, + "learning_rate": 2.8462883762873842e-06, + "loss": 0.1046, + "num_input_tokens_seen": 8819072, + "step": 32885 + }, + { + "epoch": 8.621231979030144, + "grad_norm": 1.7622873783111572, + "learning_rate": 2.8409906417906e-06, + "loss": 0.0862, + "num_input_tokens_seen": 8820272, + "step": 32890 + }, + { + "epoch": 8.62254259501966, + "grad_norm": 1.7031365633010864, + "learning_rate": 2.8356975451361873e-06, + "loss": 0.1378, + "num_input_tokens_seen": 8821520, + "step": 32895 + }, + { + "epoch": 8.623853211009175, + "grad_norm": 1.5930850505828857, + "learning_rate": 2.8304090874319843e-06, + "loss": 0.2127, + "num_input_tokens_seen": 8822768, + "step": 32900 + }, + { + "epoch": 8.625163826998689, + "grad_norm": 4.627859115600586, + "learning_rate": 2.8251252697848505e-06, + "loss": 0.1673, + "num_input_tokens_seen": 8824384, + "step": 32905 + }, + { + "epoch": 8.626474442988204, + "grad_norm": 1.7609539031982422, + "learning_rate": 2.819846093300682e-06, + "loss": 0.2669, + "num_input_tokens_seen": 8825712, + "step": 32910 + }, + { + "epoch": 8.62778505897772, + "grad_norm": 1.8368617296218872, + "learning_rate": 2.814571559084403e-06, + "loss": 0.1284, + "num_input_tokens_seen": 8826688, + "step": 32915 + }, + { + "epoch": 8.629095674967235, + "grad_norm": 1.633794903755188, + "learning_rate": 2.8093016682399636e-06, + "loss": 0.1911, + "num_input_tokens_seen": 8828144, + "step": 32920 + }, + { + "epoch": 8.63040629095675, + "grad_norm": 0.8879474401473999, + "learning_rate": 2.804036421870343e-06, + "loss": 0.1773, + "num_input_tokens_seen": 8829856, + "step": 32925 + }, + { + "epoch": 8.631716906946265, + "grad_norm": 2.6837399005889893, + "learning_rate": 2.798775821077551e-06, + "loss": 0.1039, + "num_input_tokens_seen": 8831280, + "step": 32930 + }, + { + "epoch": 8.63302752293578, + "grad_norm": 0.9350231885910034, + "learning_rate": 2.7935198669626155e-06, + "loss": 0.2078, + "num_input_tokens_seen": 8832688, + "step": 32935 + }, + { + "epoch": 8.634338138925294, + "grad_norm": 1.909218430519104, + "learning_rate": 2.788268560625601e-06, + "loss": 0.1128, + "num_input_tokens_seen": 8833952, + "step": 32940 + }, + { + "epoch": 8.63564875491481, + "grad_norm": 5.282109260559082, + "learning_rate": 2.7830219031655973e-06, + "loss": 0.1981, + "num_input_tokens_seen": 8835504, + "step": 32945 + }, + { + "epoch": 8.636959370904325, + "grad_norm": 3.257240056991577, + "learning_rate": 2.777779895680721e-06, + "loss": 0.1988, + "num_input_tokens_seen": 8837216, + "step": 32950 + }, + { + "epoch": 8.63826998689384, + "grad_norm": 2.1625518798828125, + "learning_rate": 2.772542539268114e-06, + "loss": 0.1576, + "num_input_tokens_seen": 8838176, + "step": 32955 + }, + { + "epoch": 8.639580602883354, + "grad_norm": 3.593954086303711, + "learning_rate": 2.7673098350239485e-06, + "loss": 0.1484, + "num_input_tokens_seen": 8839440, + "step": 32960 + }, + { + "epoch": 8.64089121887287, + "grad_norm": 3.5772907733917236, + "learning_rate": 2.7620817840434216e-06, + "loss": 0.1276, + "num_input_tokens_seen": 8840688, + "step": 32965 + }, + { + "epoch": 8.642201834862385, + "grad_norm": 0.7688363194465637, + "learning_rate": 2.756858387420758e-06, + "loss": 0.185, + "num_input_tokens_seen": 8841680, + "step": 32970 + }, + { + "epoch": 8.643512450851901, + "grad_norm": 1.5364524126052856, + "learning_rate": 2.7516396462491934e-06, + "loss": 0.1922, + "num_input_tokens_seen": 8843120, + "step": 32975 + }, + { + "epoch": 8.644823066841415, + "grad_norm": 2.282999277114868, + "learning_rate": 2.7464255616210106e-06, + "loss": 0.1164, + "num_input_tokens_seen": 8844752, + "step": 32980 + }, + { + "epoch": 8.64613368283093, + "grad_norm": 1.6297131776809692, + "learning_rate": 2.7412161346275057e-06, + "loss": 0.207, + "num_input_tokens_seen": 8845856, + "step": 32985 + }, + { + "epoch": 8.647444298820446, + "grad_norm": 2.6451609134674072, + "learning_rate": 2.7360113663590036e-06, + "loss": 0.2736, + "num_input_tokens_seen": 8846976, + "step": 32990 + }, + { + "epoch": 8.648754914809961, + "grad_norm": 2.074977159500122, + "learning_rate": 2.7308112579048555e-06, + "loss": 0.2388, + "num_input_tokens_seen": 8848832, + "step": 32995 + }, + { + "epoch": 8.650065530799475, + "grad_norm": 1.8765040636062622, + "learning_rate": 2.7256158103534298e-06, + "loss": 0.1528, + "num_input_tokens_seen": 8850080, + "step": 33000 + }, + { + "epoch": 8.65137614678899, + "grad_norm": 1.0637751817703247, + "learning_rate": 2.7204250247921308e-06, + "loss": 0.1565, + "num_input_tokens_seen": 8851552, + "step": 33005 + }, + { + "epoch": 8.652686762778506, + "grad_norm": 2.0576236248016357, + "learning_rate": 2.7152389023073766e-06, + "loss": 0.1936, + "num_input_tokens_seen": 8852688, + "step": 33010 + }, + { + "epoch": 8.653997378768022, + "grad_norm": 0.4348452389240265, + "learning_rate": 2.710057443984615e-06, + "loss": 0.1263, + "num_input_tokens_seen": 8854176, + "step": 33015 + }, + { + "epoch": 8.655307994757536, + "grad_norm": 3.875850200653076, + "learning_rate": 2.7048806509083146e-06, + "loss": 0.1609, + "num_input_tokens_seen": 8855584, + "step": 33020 + }, + { + "epoch": 8.656618610747051, + "grad_norm": 2.6766958236694336, + "learning_rate": 2.6997085241619728e-06, + "loss": 0.2143, + "num_input_tokens_seen": 8857056, + "step": 33025 + }, + { + "epoch": 8.657929226736567, + "grad_norm": 2.3186142444610596, + "learning_rate": 2.6945410648281044e-06, + "loss": 0.2245, + "num_input_tokens_seen": 8858304, + "step": 33030 + }, + { + "epoch": 8.65923984272608, + "grad_norm": 1.4786971807479858, + "learning_rate": 2.6893782739882523e-06, + "loss": 0.1428, + "num_input_tokens_seen": 8859936, + "step": 33035 + }, + { + "epoch": 8.660550458715596, + "grad_norm": 4.366492748260498, + "learning_rate": 2.6842201527229743e-06, + "loss": 0.1372, + "num_input_tokens_seen": 8860976, + "step": 33040 + }, + { + "epoch": 8.661861074705111, + "grad_norm": 1.1067733764648438, + "learning_rate": 2.6790667021118626e-06, + "loss": 0.0847, + "num_input_tokens_seen": 8861792, + "step": 33045 + }, + { + "epoch": 8.663171690694627, + "grad_norm": 1.9842478036880493, + "learning_rate": 2.6739179232335264e-06, + "loss": 0.3054, + "num_input_tokens_seen": 8862880, + "step": 33050 + }, + { + "epoch": 8.66448230668414, + "grad_norm": 3.034677028656006, + "learning_rate": 2.6687738171655873e-06, + "loss": 0.2312, + "num_input_tokens_seen": 8863952, + "step": 33055 + }, + { + "epoch": 8.665792922673656, + "grad_norm": 1.3077737092971802, + "learning_rate": 2.6636343849847033e-06, + "loss": 0.2357, + "num_input_tokens_seen": 8865632, + "step": 33060 + }, + { + "epoch": 8.667103538663172, + "grad_norm": 2.183093786239624, + "learning_rate": 2.6584996277665475e-06, + "loss": 0.1063, + "num_input_tokens_seen": 8867136, + "step": 33065 + }, + { + "epoch": 8.668414154652687, + "grad_norm": 2.175110340118408, + "learning_rate": 2.653369546585813e-06, + "loss": 0.1594, + "num_input_tokens_seen": 8868288, + "step": 33070 + }, + { + "epoch": 8.669724770642201, + "grad_norm": 0.9121313691139221, + "learning_rate": 2.6482441425162235e-06, + "loss": 0.098, + "num_input_tokens_seen": 8869584, + "step": 33075 + }, + { + "epoch": 8.671035386631717, + "grad_norm": 2.205355167388916, + "learning_rate": 2.6431234166305135e-06, + "loss": 0.1694, + "num_input_tokens_seen": 8871216, + "step": 33080 + }, + { + "epoch": 8.672346002621232, + "grad_norm": 0.3517792522907257, + "learning_rate": 2.6380073700004504e-06, + "loss": 0.1065, + "num_input_tokens_seen": 8872160, + "step": 33085 + }, + { + "epoch": 8.673656618610748, + "grad_norm": 2.594083547592163, + "learning_rate": 2.6328960036967996e-06, + "loss": 0.1169, + "num_input_tokens_seen": 8873792, + "step": 33090 + }, + { + "epoch": 8.674967234600262, + "grad_norm": 4.8001389503479, + "learning_rate": 2.627789318789367e-06, + "loss": 0.2496, + "num_input_tokens_seen": 8874960, + "step": 33095 + }, + { + "epoch": 8.676277850589777, + "grad_norm": 2.1198573112487793, + "learning_rate": 2.6226873163469752e-06, + "loss": 0.1837, + "num_input_tokens_seen": 8876304, + "step": 33100 + }, + { + "epoch": 8.677588466579293, + "grad_norm": 1.8333083391189575, + "learning_rate": 2.6175899974374614e-06, + "loss": 0.0649, + "num_input_tokens_seen": 8877312, + "step": 33105 + }, + { + "epoch": 8.678899082568808, + "grad_norm": 2.0024633407592773, + "learning_rate": 2.6124973631276888e-06, + "loss": 0.1572, + "num_input_tokens_seen": 8878544, + "step": 33110 + }, + { + "epoch": 8.680209698558322, + "grad_norm": 5.713492393493652, + "learning_rate": 2.607409414483536e-06, + "loss": 0.293, + "num_input_tokens_seen": 8879424, + "step": 33115 + }, + { + "epoch": 8.681520314547837, + "grad_norm": 2.249950885772705, + "learning_rate": 2.6023261525699003e-06, + "loss": 0.0961, + "num_input_tokens_seen": 8880592, + "step": 33120 + }, + { + "epoch": 8.682830930537353, + "grad_norm": 0.8176953196525574, + "learning_rate": 2.597247578450701e-06, + "loss": 0.3356, + "num_input_tokens_seen": 8882272, + "step": 33125 + }, + { + "epoch": 8.684141546526867, + "grad_norm": 2.4637606143951416, + "learning_rate": 2.592173693188876e-06, + "loss": 0.1241, + "num_input_tokens_seen": 8883248, + "step": 33130 + }, + { + "epoch": 8.685452162516382, + "grad_norm": 2.232689380645752, + "learning_rate": 2.5871044978463764e-06, + "loss": 0.1491, + "num_input_tokens_seen": 8884784, + "step": 33135 + }, + { + "epoch": 8.686762778505898, + "grad_norm": 3.4281744956970215, + "learning_rate": 2.582039993484181e-06, + "loss": 0.2578, + "num_input_tokens_seen": 8886576, + "step": 33140 + }, + { + "epoch": 8.688073394495413, + "grad_norm": 2.208387613296509, + "learning_rate": 2.576980181162278e-06, + "loss": 0.1208, + "num_input_tokens_seen": 8887616, + "step": 33145 + }, + { + "epoch": 8.689384010484927, + "grad_norm": 1.9373735189437866, + "learning_rate": 2.571925061939681e-06, + "loss": 0.1706, + "num_input_tokens_seen": 8889232, + "step": 33150 + }, + { + "epoch": 8.690694626474443, + "grad_norm": 1.442133903503418, + "learning_rate": 2.566874636874414e-06, + "loss": 0.2035, + "num_input_tokens_seen": 8890432, + "step": 33155 + }, + { + "epoch": 8.692005242463958, + "grad_norm": 5.62154483795166, + "learning_rate": 2.561828907023525e-06, + "loss": 0.2258, + "num_input_tokens_seen": 8892160, + "step": 33160 + }, + { + "epoch": 8.693315858453474, + "grad_norm": 2.029691696166992, + "learning_rate": 2.556787873443081e-06, + "loss": 0.148, + "num_input_tokens_seen": 8893360, + "step": 33165 + }, + { + "epoch": 8.694626474442988, + "grad_norm": 2.098461151123047, + "learning_rate": 2.5517515371881516e-06, + "loss": 0.1045, + "num_input_tokens_seen": 8895104, + "step": 33170 + }, + { + "epoch": 8.695937090432503, + "grad_norm": 1.5230528116226196, + "learning_rate": 2.5467198993128355e-06, + "loss": 0.0835, + "num_input_tokens_seen": 8896304, + "step": 33175 + }, + { + "epoch": 8.697247706422019, + "grad_norm": 3.9353201389312744, + "learning_rate": 2.541692960870251e-06, + "loss": 0.2342, + "num_input_tokens_seen": 8897440, + "step": 33180 + }, + { + "epoch": 8.698558322411534, + "grad_norm": 1.6850121021270752, + "learning_rate": 2.536670722912518e-06, + "loss": 0.1456, + "num_input_tokens_seen": 8898736, + "step": 33185 + }, + { + "epoch": 8.699868938401048, + "grad_norm": 1.3036980628967285, + "learning_rate": 2.5316531864907955e-06, + "loss": 0.1, + "num_input_tokens_seen": 8899632, + "step": 33190 + }, + { + "epoch": 8.701179554390563, + "grad_norm": 1.7211915254592896, + "learning_rate": 2.526640352655238e-06, + "loss": 0.1566, + "num_input_tokens_seen": 8900848, + "step": 33195 + }, + { + "epoch": 8.702490170380079, + "grad_norm": 1.960544228553772, + "learning_rate": 2.5216322224550226e-06, + "loss": 0.0523, + "num_input_tokens_seen": 8901808, + "step": 33200 + }, + { + "epoch": 8.703800786369595, + "grad_norm": 1.663207769393921, + "learning_rate": 2.51662879693835e-06, + "loss": 0.1004, + "num_input_tokens_seen": 8903008, + "step": 33205 + }, + { + "epoch": 8.705111402359108, + "grad_norm": 1.0737353563308716, + "learning_rate": 2.511630077152416e-06, + "loss": 0.2103, + "num_input_tokens_seen": 8904192, + "step": 33210 + }, + { + "epoch": 8.706422018348624, + "grad_norm": 2.0207016468048096, + "learning_rate": 2.50663606414345e-06, + "loss": 0.198, + "num_input_tokens_seen": 8905696, + "step": 33215 + }, + { + "epoch": 8.70773263433814, + "grad_norm": 0.7914749979972839, + "learning_rate": 2.501646758956691e-06, + "loss": 0.3192, + "num_input_tokens_seen": 8907440, + "step": 33220 + }, + { + "epoch": 8.709043250327653, + "grad_norm": 2.954555034637451, + "learning_rate": 2.49666216263639e-06, + "loss": 0.2182, + "num_input_tokens_seen": 8908544, + "step": 33225 + }, + { + "epoch": 8.710353866317169, + "grad_norm": 3.508357048034668, + "learning_rate": 2.491682276225818e-06, + "loss": 0.1102, + "num_input_tokens_seen": 8909744, + "step": 33230 + }, + { + "epoch": 8.711664482306684, + "grad_norm": 2.7497987747192383, + "learning_rate": 2.486707100767252e-06, + "loss": 0.2193, + "num_input_tokens_seen": 8911040, + "step": 33235 + }, + { + "epoch": 8.7129750982962, + "grad_norm": 4.630180358886719, + "learning_rate": 2.4817366373019902e-06, + "loss": 0.1208, + "num_input_tokens_seen": 8912144, + "step": 33240 + }, + { + "epoch": 8.714285714285714, + "grad_norm": 3.842024803161621, + "learning_rate": 2.4767708868703414e-06, + "loss": 0.1554, + "num_input_tokens_seen": 8913296, + "step": 33245 + }, + { + "epoch": 8.715596330275229, + "grad_norm": 1.6432117223739624, + "learning_rate": 2.4718098505116305e-06, + "loss": 0.1257, + "num_input_tokens_seen": 8914464, + "step": 33250 + }, + { + "epoch": 8.716906946264745, + "grad_norm": 2.5896809101104736, + "learning_rate": 2.4668535292641904e-06, + "loss": 0.2616, + "num_input_tokens_seen": 8916016, + "step": 33255 + }, + { + "epoch": 8.71821756225426, + "grad_norm": 2.2532010078430176, + "learning_rate": 2.4619019241653716e-06, + "loss": 0.1358, + "num_input_tokens_seen": 8917328, + "step": 33260 + }, + { + "epoch": 8.719528178243774, + "grad_norm": 6.243034839630127, + "learning_rate": 2.4569550362515403e-06, + "loss": 0.0946, + "num_input_tokens_seen": 8919232, + "step": 33265 + }, + { + "epoch": 8.72083879423329, + "grad_norm": 2.6471517086029053, + "learning_rate": 2.452012866558065e-06, + "loss": 0.1381, + "num_input_tokens_seen": 8920512, + "step": 33270 + }, + { + "epoch": 8.722149410222805, + "grad_norm": 2.886479377746582, + "learning_rate": 2.4470754161193406e-06, + "loss": 0.0961, + "num_input_tokens_seen": 8921680, + "step": 33275 + }, + { + "epoch": 8.72346002621232, + "grad_norm": 3.7762365341186523, + "learning_rate": 2.4421426859687597e-06, + "loss": 0.3055, + "num_input_tokens_seen": 8923056, + "step": 33280 + }, + { + "epoch": 8.724770642201834, + "grad_norm": 1.9678680896759033, + "learning_rate": 2.437214677138744e-06, + "loss": 0.0759, + "num_input_tokens_seen": 8923968, + "step": 33285 + }, + { + "epoch": 8.72608125819135, + "grad_norm": 1.7752375602722168, + "learning_rate": 2.432291390660707e-06, + "loss": 0.1588, + "num_input_tokens_seen": 8925248, + "step": 33290 + }, + { + "epoch": 8.727391874180865, + "grad_norm": 1.448912262916565, + "learning_rate": 2.427372827565086e-06, + "loss": 0.1407, + "num_input_tokens_seen": 8926320, + "step": 33295 + }, + { + "epoch": 8.728702490170381, + "grad_norm": 1.5499478578567505, + "learning_rate": 2.4224589888813263e-06, + "loss": 0.1452, + "num_input_tokens_seen": 8927840, + "step": 33300 + }, + { + "epoch": 8.730013106159895, + "grad_norm": 2.5575053691864014, + "learning_rate": 2.4175498756378924e-06, + "loss": 0.2224, + "num_input_tokens_seen": 8929024, + "step": 33305 + }, + { + "epoch": 8.73132372214941, + "grad_norm": 1.8915897607803345, + "learning_rate": 2.41264548886225e-06, + "loss": 0.2125, + "num_input_tokens_seen": 8930240, + "step": 33310 + }, + { + "epoch": 8.732634338138926, + "grad_norm": 1.9108837842941284, + "learning_rate": 2.407745829580882e-06, + "loss": 0.0564, + "num_input_tokens_seen": 8931280, + "step": 33315 + }, + { + "epoch": 8.73394495412844, + "grad_norm": 3.4087822437286377, + "learning_rate": 2.402850898819278e-06, + "loss": 0.142, + "num_input_tokens_seen": 8932480, + "step": 33320 + }, + { + "epoch": 8.735255570117955, + "grad_norm": 3.214944362640381, + "learning_rate": 2.397960697601931e-06, + "loss": 0.3063, + "num_input_tokens_seen": 8934048, + "step": 33325 + }, + { + "epoch": 8.73656618610747, + "grad_norm": 2.3850338459014893, + "learning_rate": 2.393075226952357e-06, + "loss": 0.2495, + "num_input_tokens_seen": 8935520, + "step": 33330 + }, + { + "epoch": 8.737876802096986, + "grad_norm": 0.9591131210327148, + "learning_rate": 2.388194487893078e-06, + "loss": 0.1862, + "num_input_tokens_seen": 8937248, + "step": 33335 + }, + { + "epoch": 8.7391874180865, + "grad_norm": 4.315678119659424, + "learning_rate": 2.3833184814456204e-06, + "loss": 0.1755, + "num_input_tokens_seen": 8938688, + "step": 33340 + }, + { + "epoch": 8.740498034076015, + "grad_norm": 2.8851206302642822, + "learning_rate": 2.378447208630527e-06, + "loss": 0.2125, + "num_input_tokens_seen": 8940176, + "step": 33345 + }, + { + "epoch": 8.741808650065531, + "grad_norm": 1.3297920227050781, + "learning_rate": 2.373580670467343e-06, + "loss": 0.2049, + "num_input_tokens_seen": 8941568, + "step": 33350 + }, + { + "epoch": 8.743119266055047, + "grad_norm": 1.9749054908752441, + "learning_rate": 2.3687188679746315e-06, + "loss": 0.2466, + "num_input_tokens_seen": 8943376, + "step": 33355 + }, + { + "epoch": 8.74442988204456, + "grad_norm": 0.1285613626241684, + "learning_rate": 2.363861802169959e-06, + "loss": 0.1297, + "num_input_tokens_seen": 8944832, + "step": 33360 + }, + { + "epoch": 8.745740498034076, + "grad_norm": 2.1388373374938965, + "learning_rate": 2.3590094740698955e-06, + "loss": 0.2185, + "num_input_tokens_seen": 8946368, + "step": 33365 + }, + { + "epoch": 8.747051114023591, + "grad_norm": 1.2216017246246338, + "learning_rate": 2.3541618846900316e-06, + "loss": 0.1225, + "num_input_tokens_seen": 8947904, + "step": 33370 + }, + { + "epoch": 8.748361730013107, + "grad_norm": 3.5306639671325684, + "learning_rate": 2.3493190350449557e-06, + "loss": 0.1979, + "num_input_tokens_seen": 8949168, + "step": 33375 + }, + { + "epoch": 8.74967234600262, + "grad_norm": 1.5940853357315063, + "learning_rate": 2.3444809261482653e-06, + "loss": 0.2495, + "num_input_tokens_seen": 8950880, + "step": 33380 + }, + { + "epoch": 8.750982961992136, + "grad_norm": 1.1162830591201782, + "learning_rate": 2.3396475590125753e-06, + "loss": 0.2513, + "num_input_tokens_seen": 8952688, + "step": 33385 + }, + { + "epoch": 8.752293577981652, + "grad_norm": 4.15873908996582, + "learning_rate": 2.334818934649494e-06, + "loss": 0.233, + "num_input_tokens_seen": 8954448, + "step": 33390 + }, + { + "epoch": 8.753604193971167, + "grad_norm": 2.640967607498169, + "learning_rate": 2.3299950540696504e-06, + "loss": 0.1647, + "num_input_tokens_seen": 8955872, + "step": 33395 + }, + { + "epoch": 8.754914809960681, + "grad_norm": 2.832542657852173, + "learning_rate": 2.325175918282674e-06, + "loss": 0.2053, + "num_input_tokens_seen": 8957840, + "step": 33400 + }, + { + "epoch": 8.756225425950197, + "grad_norm": 2.2102179527282715, + "learning_rate": 2.3203615282971934e-06, + "loss": 0.2338, + "num_input_tokens_seen": 8959216, + "step": 33405 + }, + { + "epoch": 8.757536041939712, + "grad_norm": 2.675659418106079, + "learning_rate": 2.315551885120859e-06, + "loss": 0.1626, + "num_input_tokens_seen": 8960464, + "step": 33410 + }, + { + "epoch": 8.758846657929226, + "grad_norm": 2.3125696182250977, + "learning_rate": 2.3107469897603152e-06, + "loss": 0.185, + "num_input_tokens_seen": 8961536, + "step": 33415 + }, + { + "epoch": 8.760157273918741, + "grad_norm": 4.198662281036377, + "learning_rate": 2.305946843221224e-06, + "loss": 0.0823, + "num_input_tokens_seen": 8962720, + "step": 33420 + }, + { + "epoch": 8.761467889908257, + "grad_norm": 0.0942690372467041, + "learning_rate": 2.3011514465082485e-06, + "loss": 0.1084, + "num_input_tokens_seen": 8963888, + "step": 33425 + }, + { + "epoch": 8.762778505897773, + "grad_norm": 1.3919621706008911, + "learning_rate": 2.2963608006250555e-06, + "loss": 0.0673, + "num_input_tokens_seen": 8965008, + "step": 33430 + }, + { + "epoch": 8.764089121887286, + "grad_norm": 2.3114070892333984, + "learning_rate": 2.2915749065743176e-06, + "loss": 0.193, + "num_input_tokens_seen": 8966432, + "step": 33435 + }, + { + "epoch": 8.765399737876802, + "grad_norm": 3.0446524620056152, + "learning_rate": 2.2867937653577173e-06, + "loss": 0.2119, + "num_input_tokens_seen": 8968304, + "step": 33440 + }, + { + "epoch": 8.766710353866317, + "grad_norm": 2.0906124114990234, + "learning_rate": 2.282017377975934e-06, + "loss": 0.2119, + "num_input_tokens_seen": 8969600, + "step": 33445 + }, + { + "epoch": 8.768020969855833, + "grad_norm": 2.0632944107055664, + "learning_rate": 2.2772457454286578e-06, + "loss": 0.1749, + "num_input_tokens_seen": 8970640, + "step": 33450 + }, + { + "epoch": 8.769331585845347, + "grad_norm": 1.9617260694503784, + "learning_rate": 2.2724788687145865e-06, + "loss": 0.1605, + "num_input_tokens_seen": 8971776, + "step": 33455 + }, + { + "epoch": 8.770642201834862, + "grad_norm": 1.3435953855514526, + "learning_rate": 2.2677167488314193e-06, + "loss": 0.164, + "num_input_tokens_seen": 8973136, + "step": 33460 + }, + { + "epoch": 8.771952817824378, + "grad_norm": 1.4298346042633057, + "learning_rate": 2.2629593867758564e-06, + "loss": 0.2087, + "num_input_tokens_seen": 8974544, + "step": 33465 + }, + { + "epoch": 8.773263433813893, + "grad_norm": 2.6691794395446777, + "learning_rate": 2.258206783543601e-06, + "loss": 0.1193, + "num_input_tokens_seen": 8975632, + "step": 33470 + }, + { + "epoch": 8.774574049803407, + "grad_norm": 1.1134114265441895, + "learning_rate": 2.253458940129383e-06, + "loss": 0.1578, + "num_input_tokens_seen": 8976800, + "step": 33475 + }, + { + "epoch": 8.775884665792923, + "grad_norm": 1.4686230421066284, + "learning_rate": 2.2487158575269006e-06, + "loss": 0.0597, + "num_input_tokens_seen": 8977952, + "step": 33480 + }, + { + "epoch": 8.777195281782438, + "grad_norm": 1.8059629201889038, + "learning_rate": 2.243977536728878e-06, + "loss": 0.1292, + "num_input_tokens_seen": 8979392, + "step": 33485 + }, + { + "epoch": 8.778505897771954, + "grad_norm": 1.925597071647644, + "learning_rate": 2.2392439787270373e-06, + "loss": 0.1213, + "num_input_tokens_seen": 8980416, + "step": 33490 + }, + { + "epoch": 8.779816513761467, + "grad_norm": 2.76880145072937, + "learning_rate": 2.234515184512104e-06, + "loss": 0.1381, + "num_input_tokens_seen": 8981904, + "step": 33495 + }, + { + "epoch": 8.781127129750983, + "grad_norm": 2.28420090675354, + "learning_rate": 2.229791155073807e-06, + "loss": 0.1437, + "num_input_tokens_seen": 8983696, + "step": 33500 + }, + { + "epoch": 8.782437745740499, + "grad_norm": 0.5457587838172913, + "learning_rate": 2.225071891400879e-06, + "loss": 0.1312, + "num_input_tokens_seen": 8984752, + "step": 33505 + }, + { + "epoch": 8.783748361730012, + "grad_norm": 3.449852705001831, + "learning_rate": 2.2203573944810514e-06, + "loss": 0.1104, + "num_input_tokens_seen": 8985984, + "step": 33510 + }, + { + "epoch": 8.785058977719528, + "grad_norm": 4.108770847320557, + "learning_rate": 2.21564766530106e-06, + "loss": 0.1728, + "num_input_tokens_seen": 8987232, + "step": 33515 + }, + { + "epoch": 8.786369593709043, + "grad_norm": 3.8518495559692383, + "learning_rate": 2.210942704846647e-06, + "loss": 0.177, + "num_input_tokens_seen": 8988912, + "step": 33520 + }, + { + "epoch": 8.787680209698559, + "grad_norm": 1.3189457654953003, + "learning_rate": 2.2062425141025465e-06, + "loss": 0.1181, + "num_input_tokens_seen": 8990144, + "step": 33525 + }, + { + "epoch": 8.788990825688073, + "grad_norm": 2.191422939300537, + "learning_rate": 2.2015470940525003e-06, + "loss": 0.2455, + "num_input_tokens_seen": 8991568, + "step": 33530 + }, + { + "epoch": 8.790301441677588, + "grad_norm": 4.300930023193359, + "learning_rate": 2.1968564456792573e-06, + "loss": 0.2568, + "num_input_tokens_seen": 8993072, + "step": 33535 + }, + { + "epoch": 8.791612057667104, + "grad_norm": 3.386653184890747, + "learning_rate": 2.1921705699645607e-06, + "loss": 0.1453, + "num_input_tokens_seen": 8994336, + "step": 33540 + }, + { + "epoch": 8.79292267365662, + "grad_norm": 3.562941789627075, + "learning_rate": 2.187489467889153e-06, + "loss": 0.1371, + "num_input_tokens_seen": 8995632, + "step": 33545 + }, + { + "epoch": 8.794233289646133, + "grad_norm": 1.6556742191314697, + "learning_rate": 2.182813140432785e-06, + "loss": 0.0841, + "num_input_tokens_seen": 8996704, + "step": 33550 + }, + { + "epoch": 8.795543905635649, + "grad_norm": 1.364064335823059, + "learning_rate": 2.1781415885742037e-06, + "loss": 0.1464, + "num_input_tokens_seen": 8998608, + "step": 33555 + }, + { + "epoch": 8.796854521625164, + "grad_norm": 1.9660887718200684, + "learning_rate": 2.1734748132911515e-06, + "loss": 0.1729, + "num_input_tokens_seen": 8999952, + "step": 33560 + }, + { + "epoch": 8.79816513761468, + "grad_norm": 2.608043670654297, + "learning_rate": 2.1688128155603783e-06, + "loss": 0.1045, + "num_input_tokens_seen": 9001312, + "step": 33565 + }, + { + "epoch": 8.799475753604193, + "grad_norm": 2.5513908863067627, + "learning_rate": 2.1641555963576366e-06, + "loss": 0.09, + "num_input_tokens_seen": 9002528, + "step": 33570 + }, + { + "epoch": 8.800786369593709, + "grad_norm": 1.938335657119751, + "learning_rate": 2.159503156657669e-06, + "loss": 0.1261, + "num_input_tokens_seen": 9003776, + "step": 33575 + }, + { + "epoch": 8.802096985583225, + "grad_norm": 3.64111590385437, + "learning_rate": 2.1548554974342244e-06, + "loss": 0.1387, + "num_input_tokens_seen": 9004768, + "step": 33580 + }, + { + "epoch": 8.80340760157274, + "grad_norm": 2.0862410068511963, + "learning_rate": 2.15021261966005e-06, + "loss": 0.2295, + "num_input_tokens_seen": 9005984, + "step": 33585 + }, + { + "epoch": 8.804718217562254, + "grad_norm": 2.5233590602874756, + "learning_rate": 2.145574524306898e-06, + "loss": 0.1575, + "num_input_tokens_seen": 9007200, + "step": 33590 + }, + { + "epoch": 8.80602883355177, + "grad_norm": 2.317344903945923, + "learning_rate": 2.140941212345507e-06, + "loss": 0.1767, + "num_input_tokens_seen": 9009152, + "step": 33595 + }, + { + "epoch": 8.807339449541285, + "grad_norm": 0.44534173607826233, + "learning_rate": 2.136312684745623e-06, + "loss": 0.0769, + "num_input_tokens_seen": 9011744, + "step": 33600 + }, + { + "epoch": 8.808650065530799, + "grad_norm": 1.4991098642349243, + "learning_rate": 2.1316889424759874e-06, + "loss": 0.1934, + "num_input_tokens_seen": 9013424, + "step": 33605 + }, + { + "epoch": 8.809960681520314, + "grad_norm": 5.01230001449585, + "learning_rate": 2.1270699865043405e-06, + "loss": 0.158, + "num_input_tokens_seen": 9014384, + "step": 33610 + }, + { + "epoch": 8.81127129750983, + "grad_norm": 3.1978752613067627, + "learning_rate": 2.122455817797428e-06, + "loss": 0.1958, + "num_input_tokens_seen": 9015936, + "step": 33615 + }, + { + "epoch": 8.812581913499345, + "grad_norm": 1.0260043144226074, + "learning_rate": 2.1178464373209807e-06, + "loss": 0.1525, + "num_input_tokens_seen": 9017040, + "step": 33620 + }, + { + "epoch": 8.813892529488859, + "grad_norm": 3.6953792572021484, + "learning_rate": 2.1132418460397376e-06, + "loss": 0.3731, + "num_input_tokens_seen": 9018464, + "step": 33625 + }, + { + "epoch": 8.815203145478375, + "grad_norm": 1.896102786064148, + "learning_rate": 2.1086420449174303e-06, + "loss": 0.1227, + "num_input_tokens_seen": 9019648, + "step": 33630 + }, + { + "epoch": 8.81651376146789, + "grad_norm": 1.8911864757537842, + "learning_rate": 2.1040470349167923e-06, + "loss": 0.2122, + "num_input_tokens_seen": 9021360, + "step": 33635 + }, + { + "epoch": 8.817824377457406, + "grad_norm": 2.480524778366089, + "learning_rate": 2.0994568169995455e-06, + "loss": 0.1423, + "num_input_tokens_seen": 9022496, + "step": 33640 + }, + { + "epoch": 8.81913499344692, + "grad_norm": 1.2675197124481201, + "learning_rate": 2.0948713921264214e-06, + "loss": 0.1275, + "num_input_tokens_seen": 9023792, + "step": 33645 + }, + { + "epoch": 8.820445609436435, + "grad_norm": 1.2327919006347656, + "learning_rate": 2.0902907612571355e-06, + "loss": 0.1056, + "num_input_tokens_seen": 9024928, + "step": 33650 + }, + { + "epoch": 8.82175622542595, + "grad_norm": 2.5025827884674072, + "learning_rate": 2.085714925350407e-06, + "loss": 0.2146, + "num_input_tokens_seen": 9026256, + "step": 33655 + }, + { + "epoch": 8.823066841415466, + "grad_norm": 2.05829119682312, + "learning_rate": 2.0811438853639537e-06, + "loss": 0.2702, + "num_input_tokens_seen": 9027904, + "step": 33660 + }, + { + "epoch": 8.82437745740498, + "grad_norm": 1.9694416522979736, + "learning_rate": 2.076577642254485e-06, + "loss": 0.2164, + "num_input_tokens_seen": 9029168, + "step": 33665 + }, + { + "epoch": 8.825688073394495, + "grad_norm": 1.7082953453063965, + "learning_rate": 2.072016196977708e-06, + "loss": 0.1666, + "num_input_tokens_seen": 9030336, + "step": 33670 + }, + { + "epoch": 8.82699868938401, + "grad_norm": 2.9219110012054443, + "learning_rate": 2.0674595504883265e-06, + "loss": 0.1395, + "num_input_tokens_seen": 9031488, + "step": 33675 + }, + { + "epoch": 8.828309305373526, + "grad_norm": 2.2398135662078857, + "learning_rate": 2.062907703740033e-06, + "loss": 0.1438, + "num_input_tokens_seen": 9033152, + "step": 33680 + }, + { + "epoch": 8.82961992136304, + "grad_norm": 2.529118776321411, + "learning_rate": 2.0583606576855236e-06, + "loss": 0.1094, + "num_input_tokens_seen": 9034352, + "step": 33685 + }, + { + "epoch": 8.830930537352556, + "grad_norm": 0.9645698070526123, + "learning_rate": 2.0538184132764897e-06, + "loss": 0.1835, + "num_input_tokens_seen": 9035712, + "step": 33690 + }, + { + "epoch": 8.832241153342071, + "grad_norm": 2.3424997329711914, + "learning_rate": 2.0492809714636156e-06, + "loss": 0.1343, + "num_input_tokens_seen": 9036784, + "step": 33695 + }, + { + "epoch": 8.833551769331585, + "grad_norm": 1.4144335985183716, + "learning_rate": 2.044748333196572e-06, + "loss": 0.1231, + "num_input_tokens_seen": 9038256, + "step": 33700 + }, + { + "epoch": 8.8348623853211, + "grad_norm": 1.7676888704299927, + "learning_rate": 2.040220499424042e-06, + "loss": 0.1329, + "num_input_tokens_seen": 9039712, + "step": 33705 + }, + { + "epoch": 8.836173001310616, + "grad_norm": 2.019258737564087, + "learning_rate": 2.0356974710936976e-06, + "loss": 0.1605, + "num_input_tokens_seen": 9041440, + "step": 33710 + }, + { + "epoch": 8.837483617300132, + "grad_norm": 2.06618070602417, + "learning_rate": 2.0311792491521874e-06, + "loss": 0.0897, + "num_input_tokens_seen": 9042448, + "step": 33715 + }, + { + "epoch": 8.838794233289645, + "grad_norm": 3.100057601928711, + "learning_rate": 2.0266658345451715e-06, + "loss": 0.1686, + "num_input_tokens_seen": 9043824, + "step": 33720 + }, + { + "epoch": 8.840104849279161, + "grad_norm": 2.610548973083496, + "learning_rate": 2.022157228217303e-06, + "loss": 0.1673, + "num_input_tokens_seen": 9045168, + "step": 33725 + }, + { + "epoch": 8.841415465268676, + "grad_norm": 2.6827163696289062, + "learning_rate": 2.0176534311122263e-06, + "loss": 0.1203, + "num_input_tokens_seen": 9046368, + "step": 33730 + }, + { + "epoch": 8.842726081258192, + "grad_norm": 4.10058069229126, + "learning_rate": 2.0131544441725715e-06, + "loss": 0.221, + "num_input_tokens_seen": 9047520, + "step": 33735 + }, + { + "epoch": 8.844036697247706, + "grad_norm": 3.920907974243164, + "learning_rate": 2.0086602683399764e-06, + "loss": 0.1466, + "num_input_tokens_seen": 9048960, + "step": 33740 + }, + { + "epoch": 8.845347313237221, + "grad_norm": 1.9525649547576904, + "learning_rate": 2.0041709045550612e-06, + "loss": 0.0581, + "num_input_tokens_seen": 9049744, + "step": 33745 + }, + { + "epoch": 8.846657929226737, + "grad_norm": 3.1506361961364746, + "learning_rate": 1.9996863537574406e-06, + "loss": 0.2987, + "num_input_tokens_seen": 9051040, + "step": 33750 + }, + { + "epoch": 8.847968545216252, + "grad_norm": 4.543989658355713, + "learning_rate": 1.9952066168857252e-06, + "loss": 0.0882, + "num_input_tokens_seen": 9051920, + "step": 33755 + }, + { + "epoch": 8.849279161205766, + "grad_norm": 5.3707194328308105, + "learning_rate": 1.990731694877515e-06, + "loss": 0.5421, + "num_input_tokens_seen": 9053600, + "step": 33760 + }, + { + "epoch": 8.850589777195282, + "grad_norm": 2.8904550075531006, + "learning_rate": 1.9862615886694054e-06, + "loss": 0.1842, + "num_input_tokens_seen": 9054848, + "step": 33765 + }, + { + "epoch": 8.851900393184797, + "grad_norm": 3.107126235961914, + "learning_rate": 1.9817962991969805e-06, + "loss": 0.1591, + "num_input_tokens_seen": 9056032, + "step": 33770 + }, + { + "epoch": 8.853211009174313, + "grad_norm": 3.347299337387085, + "learning_rate": 1.977335827394816e-06, + "loss": 0.1981, + "num_input_tokens_seen": 9057216, + "step": 33775 + }, + { + "epoch": 8.854521625163827, + "grad_norm": 2.5165531635284424, + "learning_rate": 1.972880174196484e-06, + "loss": 0.157, + "num_input_tokens_seen": 9058528, + "step": 33780 + }, + { + "epoch": 8.855832241153342, + "grad_norm": 3.334437131881714, + "learning_rate": 1.9684293405345445e-06, + "loss": 0.143, + "num_input_tokens_seen": 9059568, + "step": 33785 + }, + { + "epoch": 8.857142857142858, + "grad_norm": 7.18773889541626, + "learning_rate": 1.9639833273405516e-06, + "loss": 0.2382, + "num_input_tokens_seen": 9060960, + "step": 33790 + }, + { + "epoch": 8.858453473132371, + "grad_norm": 2.2333734035491943, + "learning_rate": 1.959542135545042e-06, + "loss": 0.1327, + "num_input_tokens_seen": 9062576, + "step": 33795 + }, + { + "epoch": 8.859764089121887, + "grad_norm": 0.33408164978027344, + "learning_rate": 1.9551057660775527e-06, + "loss": 0.1541, + "num_input_tokens_seen": 9064096, + "step": 33800 + }, + { + "epoch": 8.861074705111402, + "grad_norm": 1.498545527458191, + "learning_rate": 1.950674219866608e-06, + "loss": 0.2036, + "num_input_tokens_seen": 9065552, + "step": 33805 + }, + { + "epoch": 8.862385321100918, + "grad_norm": 2.102759599685669, + "learning_rate": 1.9462474978397204e-06, + "loss": 0.1349, + "num_input_tokens_seen": 9066992, + "step": 33810 + }, + { + "epoch": 8.863695937090432, + "grad_norm": 3.1362037658691406, + "learning_rate": 1.941825600923397e-06, + "loss": 0.195, + "num_input_tokens_seen": 9068368, + "step": 33815 + }, + { + "epoch": 8.865006553079947, + "grad_norm": 3.523073673248291, + "learning_rate": 1.937408530043136e-06, + "loss": 0.3037, + "num_input_tokens_seen": 9069840, + "step": 33820 + }, + { + "epoch": 8.866317169069463, + "grad_norm": 2.2178852558135986, + "learning_rate": 1.9329962861234257e-06, + "loss": 0.2511, + "num_input_tokens_seen": 9071248, + "step": 33825 + }, + { + "epoch": 8.867627785058978, + "grad_norm": 1.3822157382965088, + "learning_rate": 1.9285888700877303e-06, + "loss": 0.1433, + "num_input_tokens_seen": 9072800, + "step": 33830 + }, + { + "epoch": 8.868938401048492, + "grad_norm": 3.1073691844940186, + "learning_rate": 1.92418628285852e-06, + "loss": 0.1834, + "num_input_tokens_seen": 9073904, + "step": 33835 + }, + { + "epoch": 8.870249017038008, + "grad_norm": 1.1330821514129639, + "learning_rate": 1.9197885253572497e-06, + "loss": 0.0932, + "num_input_tokens_seen": 9075184, + "step": 33840 + }, + { + "epoch": 8.871559633027523, + "grad_norm": 1.6485708951950073, + "learning_rate": 1.915395598504363e-06, + "loss": 0.1363, + "num_input_tokens_seen": 9076464, + "step": 33845 + }, + { + "epoch": 8.872870249017039, + "grad_norm": 2.272101402282715, + "learning_rate": 1.9110075032192894e-06, + "loss": 0.2461, + "num_input_tokens_seen": 9077680, + "step": 33850 + }, + { + "epoch": 8.874180865006553, + "grad_norm": 0.9759554862976074, + "learning_rate": 1.9066242404204487e-06, + "loss": 0.0684, + "num_input_tokens_seen": 9078976, + "step": 33855 + }, + { + "epoch": 8.875491480996068, + "grad_norm": 1.6252402067184448, + "learning_rate": 1.902245811025255e-06, + "loss": 0.1384, + "num_input_tokens_seen": 9080352, + "step": 33860 + }, + { + "epoch": 8.876802096985584, + "grad_norm": 1.5483505725860596, + "learning_rate": 1.8978722159501029e-06, + "loss": 0.168, + "num_input_tokens_seen": 9081552, + "step": 33865 + }, + { + "epoch": 8.8781127129751, + "grad_norm": 1.3528289794921875, + "learning_rate": 1.8935034561103764e-06, + "loss": 0.1477, + "num_input_tokens_seen": 9082752, + "step": 33870 + }, + { + "epoch": 8.879423328964613, + "grad_norm": 3.6910901069641113, + "learning_rate": 1.889139532420453e-06, + "loss": 0.0882, + "num_input_tokens_seen": 9083936, + "step": 33875 + }, + { + "epoch": 8.880733944954128, + "grad_norm": 4.79214334487915, + "learning_rate": 1.8847804457936934e-06, + "loss": 0.082, + "num_input_tokens_seen": 9085264, + "step": 33880 + }, + { + "epoch": 8.882044560943644, + "grad_norm": 1.4000964164733887, + "learning_rate": 1.8804261971424458e-06, + "loss": 0.0498, + "num_input_tokens_seen": 9086512, + "step": 33885 + }, + { + "epoch": 8.883355176933158, + "grad_norm": 2.6986255645751953, + "learning_rate": 1.8760767873780476e-06, + "loss": 0.0785, + "num_input_tokens_seen": 9087648, + "step": 33890 + }, + { + "epoch": 8.884665792922673, + "grad_norm": 1.8135570287704468, + "learning_rate": 1.871732217410821e-06, + "loss": 0.2472, + "num_input_tokens_seen": 9089120, + "step": 33895 + }, + { + "epoch": 8.885976408912189, + "grad_norm": 0.7207011580467224, + "learning_rate": 1.8673924881500826e-06, + "loss": 0.0793, + "num_input_tokens_seen": 9090368, + "step": 33900 + }, + { + "epoch": 8.887287024901704, + "grad_norm": 1.768234133720398, + "learning_rate": 1.8630576005041228e-06, + "loss": 0.1235, + "num_input_tokens_seen": 9091552, + "step": 33905 + }, + { + "epoch": 8.888597640891218, + "grad_norm": 2.4423093795776367, + "learning_rate": 1.8587275553802352e-06, + "loss": 0.1123, + "num_input_tokens_seen": 9092800, + "step": 33910 + }, + { + "epoch": 8.889908256880734, + "grad_norm": 2.616999626159668, + "learning_rate": 1.854402353684681e-06, + "loss": 0.2486, + "num_input_tokens_seen": 9094512, + "step": 33915 + }, + { + "epoch": 8.89121887287025, + "grad_norm": 0.03867405652999878, + "learning_rate": 1.8500819963227223e-06, + "loss": 0.132, + "num_input_tokens_seen": 9095776, + "step": 33920 + }, + { + "epoch": 8.892529488859765, + "grad_norm": 3.56152081489563, + "learning_rate": 1.8457664841986023e-06, + "loss": 0.1516, + "num_input_tokens_seen": 9096992, + "step": 33925 + }, + { + "epoch": 8.893840104849279, + "grad_norm": 1.880817174911499, + "learning_rate": 1.8414558182155456e-06, + "loss": 0.0858, + "num_input_tokens_seen": 9098160, + "step": 33930 + }, + { + "epoch": 8.895150720838794, + "grad_norm": 2.577192783355713, + "learning_rate": 1.8371499992757723e-06, + "loss": 0.136, + "num_input_tokens_seen": 9100016, + "step": 33935 + }, + { + "epoch": 8.89646133682831, + "grad_norm": 2.282609224319458, + "learning_rate": 1.832849028280484e-06, + "loss": 0.1079, + "num_input_tokens_seen": 9101248, + "step": 33940 + }, + { + "epoch": 8.897771952817825, + "grad_norm": 1.4546340703964233, + "learning_rate": 1.8285529061298684e-06, + "loss": 0.0433, + "num_input_tokens_seen": 9102448, + "step": 33945 + }, + { + "epoch": 8.899082568807339, + "grad_norm": 1.8007354736328125, + "learning_rate": 1.82426163372309e-06, + "loss": 0.0644, + "num_input_tokens_seen": 9103568, + "step": 33950 + }, + { + "epoch": 8.900393184796854, + "grad_norm": 1.5881421566009521, + "learning_rate": 1.8199752119583052e-06, + "loss": 0.074, + "num_input_tokens_seen": 9104656, + "step": 33955 + }, + { + "epoch": 8.90170380078637, + "grad_norm": 1.7195794582366943, + "learning_rate": 1.8156936417326576e-06, + "loss": 0.2684, + "num_input_tokens_seen": 9106288, + "step": 33960 + }, + { + "epoch": 8.903014416775886, + "grad_norm": 1.6360911130905151, + "learning_rate": 1.8114169239422691e-06, + "loss": 0.13, + "num_input_tokens_seen": 9107728, + "step": 33965 + }, + { + "epoch": 8.9043250327654, + "grad_norm": 1.948323130607605, + "learning_rate": 1.8071450594822543e-06, + "loss": 0.1687, + "num_input_tokens_seen": 9108912, + "step": 33970 + }, + { + "epoch": 8.905635648754915, + "grad_norm": 0.7709145545959473, + "learning_rate": 1.8028780492467007e-06, + "loss": 0.0923, + "num_input_tokens_seen": 9110784, + "step": 33975 + }, + { + "epoch": 8.90694626474443, + "grad_norm": 4.7171220779418945, + "learning_rate": 1.798615894128694e-06, + "loss": 0.2098, + "num_input_tokens_seen": 9112016, + "step": 33980 + }, + { + "epoch": 8.908256880733944, + "grad_norm": 3.4476373195648193, + "learning_rate": 1.7943585950202902e-06, + "loss": 0.0852, + "num_input_tokens_seen": 9113184, + "step": 33985 + }, + { + "epoch": 8.90956749672346, + "grad_norm": 3.576739549636841, + "learning_rate": 1.7901061528125347e-06, + "loss": 0.1739, + "num_input_tokens_seen": 9114832, + "step": 33990 + }, + { + "epoch": 8.910878112712975, + "grad_norm": 1.278566837310791, + "learning_rate": 1.7858585683954597e-06, + "loss": 0.1433, + "num_input_tokens_seen": 9116128, + "step": 33995 + }, + { + "epoch": 8.91218872870249, + "grad_norm": 1.5097724199295044, + "learning_rate": 1.7816158426580742e-06, + "loss": 0.1198, + "num_input_tokens_seen": 9117408, + "step": 34000 + }, + { + "epoch": 8.913499344692005, + "grad_norm": 3.127937078475952, + "learning_rate": 1.7773779764883758e-06, + "loss": 0.0869, + "num_input_tokens_seen": 9120192, + "step": 34005 + }, + { + "epoch": 8.91480996068152, + "grad_norm": 5.336637496948242, + "learning_rate": 1.7731449707733412e-06, + "loss": 0.1751, + "num_input_tokens_seen": 9121616, + "step": 34010 + }, + { + "epoch": 8.916120576671036, + "grad_norm": 22.844207763671875, + "learning_rate": 1.7689168263989309e-06, + "loss": 0.1349, + "num_input_tokens_seen": 9122912, + "step": 34015 + }, + { + "epoch": 8.917431192660551, + "grad_norm": 4.36038064956665, + "learning_rate": 1.7646935442500872e-06, + "loss": 0.2221, + "num_input_tokens_seen": 9124336, + "step": 34020 + }, + { + "epoch": 8.918741808650065, + "grad_norm": 2.3046538829803467, + "learning_rate": 1.760475125210745e-06, + "loss": 0.0936, + "num_input_tokens_seen": 9125472, + "step": 34025 + }, + { + "epoch": 8.92005242463958, + "grad_norm": 1.380446195602417, + "learning_rate": 1.7562615701637975e-06, + "loss": 0.2667, + "num_input_tokens_seen": 9126560, + "step": 34030 + }, + { + "epoch": 8.921363040629096, + "grad_norm": 2.585806369781494, + "learning_rate": 1.7520528799911395e-06, + "loss": 0.1336, + "num_input_tokens_seen": 9127776, + "step": 34035 + }, + { + "epoch": 8.922673656618612, + "grad_norm": 1.709696650505066, + "learning_rate": 1.7478490555736471e-06, + "loss": 0.1094, + "num_input_tokens_seen": 9129248, + "step": 34040 + }, + { + "epoch": 8.923984272608125, + "grad_norm": 2.545614004135132, + "learning_rate": 1.7436500977911663e-06, + "loss": 0.2186, + "num_input_tokens_seen": 9130896, + "step": 34045 + }, + { + "epoch": 8.92529488859764, + "grad_norm": 3.4200315475463867, + "learning_rate": 1.7394560075225414e-06, + "loss": 0.1392, + "num_input_tokens_seen": 9131936, + "step": 34050 + }, + { + "epoch": 8.926605504587156, + "grad_norm": 1.5069165229797363, + "learning_rate": 1.7352667856455812e-06, + "loss": 0.1339, + "num_input_tokens_seen": 9133168, + "step": 34055 + }, + { + "epoch": 8.927916120576672, + "grad_norm": 3.2709720134735107, + "learning_rate": 1.7310824330370873e-06, + "loss": 0.1152, + "num_input_tokens_seen": 9134272, + "step": 34060 + }, + { + "epoch": 8.929226736566186, + "grad_norm": 1.815923810005188, + "learning_rate": 1.7269029505728313e-06, + "loss": 0.149, + "num_input_tokens_seen": 9135520, + "step": 34065 + }, + { + "epoch": 8.930537352555701, + "grad_norm": 6.288105487823486, + "learning_rate": 1.7227283391275772e-06, + "loss": 0.1047, + "num_input_tokens_seen": 9136752, + "step": 34070 + }, + { + "epoch": 8.931847968545217, + "grad_norm": 1.522930383682251, + "learning_rate": 1.7185585995750598e-06, + "loss": 0.0848, + "num_input_tokens_seen": 9137568, + "step": 34075 + }, + { + "epoch": 8.93315858453473, + "grad_norm": 2.559278964996338, + "learning_rate": 1.714393732788e-06, + "loss": 0.097, + "num_input_tokens_seen": 9138688, + "step": 34080 + }, + { + "epoch": 8.934469200524246, + "grad_norm": 3.984891891479492, + "learning_rate": 1.710233739638098e-06, + "loss": 0.1126, + "num_input_tokens_seen": 9139872, + "step": 34085 + }, + { + "epoch": 8.935779816513762, + "grad_norm": 2.188826322555542, + "learning_rate": 1.706078620996035e-06, + "loss": 0.2141, + "num_input_tokens_seen": 9142192, + "step": 34090 + }, + { + "epoch": 8.937090432503277, + "grad_norm": 1.2253555059432983, + "learning_rate": 1.7019283777314653e-06, + "loss": 0.2187, + "num_input_tokens_seen": 9143664, + "step": 34095 + }, + { + "epoch": 8.938401048492791, + "grad_norm": 2.3645455837249756, + "learning_rate": 1.6977830107130305e-06, + "loss": 0.1924, + "num_input_tokens_seen": 9145136, + "step": 34100 + }, + { + "epoch": 8.939711664482306, + "grad_norm": 1.5088859796524048, + "learning_rate": 1.6936425208083473e-06, + "loss": 0.1488, + "num_input_tokens_seen": 9146544, + "step": 34105 + }, + { + "epoch": 8.941022280471822, + "grad_norm": 2.19814133644104, + "learning_rate": 1.6895069088840144e-06, + "loss": 0.1278, + "num_input_tokens_seen": 9148336, + "step": 34110 + }, + { + "epoch": 8.942332896461338, + "grad_norm": 2.4326581954956055, + "learning_rate": 1.6853761758056114e-06, + "loss": 0.3652, + "num_input_tokens_seen": 9151376, + "step": 34115 + }, + { + "epoch": 8.943643512450851, + "grad_norm": 1.5332634449005127, + "learning_rate": 1.6812503224376857e-06, + "loss": 0.1084, + "num_input_tokens_seen": 9152304, + "step": 34120 + }, + { + "epoch": 8.944954128440367, + "grad_norm": 1.8397173881530762, + "learning_rate": 1.67712934964378e-06, + "loss": 0.2172, + "num_input_tokens_seen": 9153632, + "step": 34125 + }, + { + "epoch": 8.946264744429882, + "grad_norm": 2.2818446159362793, + "learning_rate": 1.673013258286399e-06, + "loss": 0.1976, + "num_input_tokens_seen": 9154928, + "step": 34130 + }, + { + "epoch": 8.947575360419398, + "grad_norm": 1.559545874595642, + "learning_rate": 1.668902049227039e-06, + "loss": 0.0852, + "num_input_tokens_seen": 9156096, + "step": 34135 + }, + { + "epoch": 8.948885976408912, + "grad_norm": 3.24996280670166, + "learning_rate": 1.664795723326168e-06, + "loss": 0.1685, + "num_input_tokens_seen": 9157472, + "step": 34140 + }, + { + "epoch": 8.950196592398427, + "grad_norm": 2.132143974304199, + "learning_rate": 1.660694281443234e-06, + "loss": 0.2282, + "num_input_tokens_seen": 9159088, + "step": 34145 + }, + { + "epoch": 8.951507208387943, + "grad_norm": 3.6784238815307617, + "learning_rate": 1.6565977244366564e-06, + "loss": 0.2231, + "num_input_tokens_seen": 9160304, + "step": 34150 + }, + { + "epoch": 8.952817824377458, + "grad_norm": 1.992252230644226, + "learning_rate": 1.6525060531638408e-06, + "loss": 0.1491, + "num_input_tokens_seen": 9161824, + "step": 34155 + }, + { + "epoch": 8.954128440366972, + "grad_norm": 1.4702649116516113, + "learning_rate": 1.6484192684811634e-06, + "loss": 0.2412, + "num_input_tokens_seen": 9163040, + "step": 34160 + }, + { + "epoch": 8.955439056356488, + "grad_norm": 0.6852478384971619, + "learning_rate": 1.644337371243987e-06, + "loss": 0.1549, + "num_input_tokens_seen": 9164352, + "step": 34165 + }, + { + "epoch": 8.956749672346003, + "grad_norm": 0.30805474519729614, + "learning_rate": 1.6402603623066448e-06, + "loss": 0.1133, + "num_input_tokens_seen": 9165488, + "step": 34170 + }, + { + "epoch": 8.958060288335517, + "grad_norm": 1.708587408065796, + "learning_rate": 1.6361882425224461e-06, + "loss": 0.1876, + "num_input_tokens_seen": 9167280, + "step": 34175 + }, + { + "epoch": 8.959370904325032, + "grad_norm": 2.0068135261535645, + "learning_rate": 1.6321210127436814e-06, + "loss": 0.1396, + "num_input_tokens_seen": 9168736, + "step": 34180 + }, + { + "epoch": 8.960681520314548, + "grad_norm": 1.8803205490112305, + "learning_rate": 1.6280586738216114e-06, + "loss": 0.1451, + "num_input_tokens_seen": 9169776, + "step": 34185 + }, + { + "epoch": 8.961992136304064, + "grad_norm": 5.143762111663818, + "learning_rate": 1.6240012266064752e-06, + "loss": 0.2845, + "num_input_tokens_seen": 9170752, + "step": 34190 + }, + { + "epoch": 8.963302752293577, + "grad_norm": 2.787557601928711, + "learning_rate": 1.619948671947491e-06, + "loss": 0.1057, + "num_input_tokens_seen": 9171984, + "step": 34195 + }, + { + "epoch": 8.964613368283093, + "grad_norm": 2.466067314147949, + "learning_rate": 1.6159010106928523e-06, + "loss": 0.1804, + "num_input_tokens_seen": 9173440, + "step": 34200 + }, + { + "epoch": 8.965923984272608, + "grad_norm": 2.806365489959717, + "learning_rate": 1.611858243689729e-06, + "loss": 0.1278, + "num_input_tokens_seen": 9174464, + "step": 34205 + }, + { + "epoch": 8.967234600262124, + "grad_norm": 1.3318612575531006, + "learning_rate": 1.6078203717842633e-06, + "loss": 0.2535, + "num_input_tokens_seen": 9176080, + "step": 34210 + }, + { + "epoch": 8.968545216251638, + "grad_norm": 1.4034357070922852, + "learning_rate": 1.6037873958215738e-06, + "loss": 0.1606, + "num_input_tokens_seen": 9178208, + "step": 34215 + }, + { + "epoch": 8.969855832241153, + "grad_norm": 1.5428791046142578, + "learning_rate": 1.5997593166457575e-06, + "loss": 0.1569, + "num_input_tokens_seen": 9179456, + "step": 34220 + }, + { + "epoch": 8.971166448230669, + "grad_norm": 1.679109811782837, + "learning_rate": 1.5957361350998846e-06, + "loss": 0.1379, + "num_input_tokens_seen": 9180736, + "step": 34225 + }, + { + "epoch": 8.972477064220184, + "grad_norm": 2.3552422523498535, + "learning_rate": 1.5917178520259979e-06, + "loss": 0.1614, + "num_input_tokens_seen": 9182416, + "step": 34230 + }, + { + "epoch": 8.973787680209698, + "grad_norm": 1.5813195705413818, + "learning_rate": 1.5877044682651194e-06, + "loss": 0.1858, + "num_input_tokens_seen": 9183904, + "step": 34235 + }, + { + "epoch": 8.975098296199214, + "grad_norm": 1.8149049282073975, + "learning_rate": 1.5836959846572408e-06, + "loss": 0.2806, + "num_input_tokens_seen": 9185232, + "step": 34240 + }, + { + "epoch": 8.97640891218873, + "grad_norm": 1.9097124338150024, + "learning_rate": 1.579692402041333e-06, + "loss": 0.1296, + "num_input_tokens_seen": 9186480, + "step": 34245 + }, + { + "epoch": 8.977719528178245, + "grad_norm": 2.2044458389282227, + "learning_rate": 1.5756937212553391e-06, + "loss": 0.1603, + "num_input_tokens_seen": 9187696, + "step": 34250 + }, + { + "epoch": 8.979030144167758, + "grad_norm": 1.3683031797409058, + "learning_rate": 1.5716999431361762e-06, + "loss": 0.1569, + "num_input_tokens_seen": 9189168, + "step": 34255 + }, + { + "epoch": 8.980340760157274, + "grad_norm": 1.0785053968429565, + "learning_rate": 1.5677110685197366e-06, + "loss": 0.0895, + "num_input_tokens_seen": 9190192, + "step": 34260 + }, + { + "epoch": 8.98165137614679, + "grad_norm": 2.627453565597534, + "learning_rate": 1.5637270982408775e-06, + "loss": 0.2296, + "num_input_tokens_seen": 9191712, + "step": 34265 + }, + { + "epoch": 8.982961992136303, + "grad_norm": 3.2021186351776123, + "learning_rate": 1.559748033133443e-06, + "loss": 0.1993, + "num_input_tokens_seen": 9192944, + "step": 34270 + }, + { + "epoch": 8.984272608125819, + "grad_norm": 1.8080775737762451, + "learning_rate": 1.555773874030242e-06, + "loss": 0.077, + "num_input_tokens_seen": 9193888, + "step": 34275 + }, + { + "epoch": 8.985583224115334, + "grad_norm": 2.7623119354248047, + "learning_rate": 1.5518046217630615e-06, + "loss": 0.1718, + "num_input_tokens_seen": 9195280, + "step": 34280 + }, + { + "epoch": 8.98689384010485, + "grad_norm": 0.3606773912906647, + "learning_rate": 1.5478402771626599e-06, + "loss": 0.1673, + "num_input_tokens_seen": 9196336, + "step": 34285 + }, + { + "epoch": 8.988204456094364, + "grad_norm": 2.452829360961914, + "learning_rate": 1.5438808410587646e-06, + "loss": 0.1574, + "num_input_tokens_seen": 9197920, + "step": 34290 + }, + { + "epoch": 8.98951507208388, + "grad_norm": 2.0748891830444336, + "learning_rate": 1.5399263142800825e-06, + "loss": 0.1699, + "num_input_tokens_seen": 9199168, + "step": 34295 + }, + { + "epoch": 8.990825688073395, + "grad_norm": 5.251930236816406, + "learning_rate": 1.5359766976542873e-06, + "loss": 0.1298, + "num_input_tokens_seen": 9200176, + "step": 34300 + }, + { + "epoch": 8.99213630406291, + "grad_norm": 3.3354709148406982, + "learning_rate": 1.5320319920080262e-06, + "loss": 0.1664, + "num_input_tokens_seen": 9201248, + "step": 34305 + }, + { + "epoch": 8.993446920052424, + "grad_norm": 1.7649786472320557, + "learning_rate": 1.5280921981669166e-06, + "loss": 0.0841, + "num_input_tokens_seen": 9202080, + "step": 34310 + }, + { + "epoch": 8.99475753604194, + "grad_norm": 1.2227129936218262, + "learning_rate": 1.5241573169555573e-06, + "loss": 0.2549, + "num_input_tokens_seen": 9203952, + "step": 34315 + }, + { + "epoch": 8.996068152031455, + "grad_norm": 1.3026989698410034, + "learning_rate": 1.5202273491975061e-06, + "loss": 0.3317, + "num_input_tokens_seen": 9204992, + "step": 34320 + }, + { + "epoch": 8.997378768020969, + "grad_norm": 2.071805477142334, + "learning_rate": 1.5163022957152994e-06, + "loss": 0.118, + "num_input_tokens_seen": 9206160, + "step": 34325 + }, + { + "epoch": 8.998689384010484, + "grad_norm": 2.2681708335876465, + "learning_rate": 1.5123821573304442e-06, + "loss": 0.1719, + "num_input_tokens_seen": 9207520, + "step": 34330 + }, + { + "epoch": 9.0, + "grad_norm": 6.18679666519165, + "learning_rate": 1.5084669348634255e-06, + "loss": 0.1442, + "num_input_tokens_seen": 9208488, + "step": 34335 + }, + { + "epoch": 9.001310615989516, + "grad_norm": 5.250423908233643, + "learning_rate": 1.5045566291336854e-06, + "loss": 0.1284, + "num_input_tokens_seen": 9209608, + "step": 34340 + }, + { + "epoch": 9.002359108781127, + "eval_loss": 0.8331424593925476, + "eval_runtime": 16.8653, + "eval_samples_per_second": 50.281, + "eval_steps_per_second": 25.14, + "num_input_tokens_seen": 9210648, + "step": 34344 + }, + { + "epoch": 9.00262123197903, + "grad_norm": 2.64365553855896, + "learning_rate": 1.500651240959644e-06, + "loss": 0.231, + "num_input_tokens_seen": 9210792, + "step": 34345 + }, + { + "epoch": 9.003931847968545, + "grad_norm": 1.4742451906204224, + "learning_rate": 1.4967507711586976e-06, + "loss": 0.0802, + "num_input_tokens_seen": 9211864, + "step": 34350 + }, + { + "epoch": 9.00524246395806, + "grad_norm": 3.101304531097412, + "learning_rate": 1.4928552205472012e-06, + "loss": 0.1898, + "num_input_tokens_seen": 9213288, + "step": 34355 + }, + { + "epoch": 9.006553079947576, + "grad_norm": 1.310290813446045, + "learning_rate": 1.488964589940489e-06, + "loss": 0.242, + "num_input_tokens_seen": 9215224, + "step": 34360 + }, + { + "epoch": 9.00786369593709, + "grad_norm": 2.2885046005249023, + "learning_rate": 1.4850788801528653e-06, + "loss": 0.1123, + "num_input_tokens_seen": 9216280, + "step": 34365 + }, + { + "epoch": 9.009174311926605, + "grad_norm": 0.42139026522636414, + "learning_rate": 1.4811980919976043e-06, + "loss": 0.0696, + "num_input_tokens_seen": 9217608, + "step": 34370 + }, + { + "epoch": 9.01048492791612, + "grad_norm": 2.106635332107544, + "learning_rate": 1.4773222262869423e-06, + "loss": 0.1911, + "num_input_tokens_seen": 9219000, + "step": 34375 + }, + { + "epoch": 9.011795543905636, + "grad_norm": 2.456885814666748, + "learning_rate": 1.4734512838320974e-06, + "loss": 0.1052, + "num_input_tokens_seen": 9220488, + "step": 34380 + }, + { + "epoch": 9.01310615989515, + "grad_norm": 1.9940434694290161, + "learning_rate": 1.4695852654432463e-06, + "loss": 0.1364, + "num_input_tokens_seen": 9221640, + "step": 34385 + }, + { + "epoch": 9.014416775884666, + "grad_norm": 0.18796886503696442, + "learning_rate": 1.4657241719295362e-06, + "loss": 0.1565, + "num_input_tokens_seen": 9223064, + "step": 34390 + }, + { + "epoch": 9.015727391874181, + "grad_norm": 1.3753341436386108, + "learning_rate": 1.4618680040990983e-06, + "loss": 0.1609, + "num_input_tokens_seen": 9224184, + "step": 34395 + }, + { + "epoch": 9.017038007863697, + "grad_norm": 1.8662570714950562, + "learning_rate": 1.4580167627590153e-06, + "loss": 0.1754, + "num_input_tokens_seen": 9225816, + "step": 34400 + }, + { + "epoch": 9.01834862385321, + "grad_norm": 0.9485692381858826, + "learning_rate": 1.4541704487153474e-06, + "loss": 0.0781, + "num_input_tokens_seen": 9226888, + "step": 34405 + }, + { + "epoch": 9.019659239842726, + "grad_norm": 1.2452952861785889, + "learning_rate": 1.4503290627731175e-06, + "loss": 0.18, + "num_input_tokens_seen": 9229096, + "step": 34410 + }, + { + "epoch": 9.020969855832242, + "grad_norm": 5.518567085266113, + "learning_rate": 1.4464926057363298e-06, + "loss": 0.1263, + "num_input_tokens_seen": 9229896, + "step": 34415 + }, + { + "epoch": 9.022280471821757, + "grad_norm": 2.4641575813293457, + "learning_rate": 1.4426610784079391e-06, + "loss": 0.1241, + "num_input_tokens_seen": 9230872, + "step": 34420 + }, + { + "epoch": 9.02359108781127, + "grad_norm": 1.2971628904342651, + "learning_rate": 1.438834481589882e-06, + "loss": 0.1738, + "num_input_tokens_seen": 9232296, + "step": 34425 + }, + { + "epoch": 9.024901703800786, + "grad_norm": 1.9030488729476929, + "learning_rate": 1.4350128160830562e-06, + "loss": 0.159, + "num_input_tokens_seen": 9233880, + "step": 34430 + }, + { + "epoch": 9.026212319790302, + "grad_norm": 1.5402261018753052, + "learning_rate": 1.4311960826873305e-06, + "loss": 0.1874, + "num_input_tokens_seen": 9235000, + "step": 34435 + }, + { + "epoch": 9.027522935779816, + "grad_norm": 2.3852765560150146, + "learning_rate": 1.427384282201541e-06, + "loss": 0.1304, + "num_input_tokens_seen": 9236232, + "step": 34440 + }, + { + "epoch": 9.028833551769331, + "grad_norm": 1.609797716140747, + "learning_rate": 1.4235774154234855e-06, + "loss": 0.1504, + "num_input_tokens_seen": 9237656, + "step": 34445 + }, + { + "epoch": 9.030144167758847, + "grad_norm": 1.6927604675292969, + "learning_rate": 1.4197754831499488e-06, + "loss": 0.1436, + "num_input_tokens_seen": 9238984, + "step": 34450 + }, + { + "epoch": 9.031454783748362, + "grad_norm": 1.0361075401306152, + "learning_rate": 1.4159784861766584e-06, + "loss": 0.1083, + "num_input_tokens_seen": 9240024, + "step": 34455 + }, + { + "epoch": 9.032765399737876, + "grad_norm": 1.4384706020355225, + "learning_rate": 1.4121864252983174e-06, + "loss": 0.1469, + "num_input_tokens_seen": 9241416, + "step": 34460 + }, + { + "epoch": 9.034076015727392, + "grad_norm": 1.5162047147750854, + "learning_rate": 1.408399301308605e-06, + "loss": 0.1163, + "num_input_tokens_seen": 9242712, + "step": 34465 + }, + { + "epoch": 9.035386631716907, + "grad_norm": 1.2983700037002563, + "learning_rate": 1.4046171150001508e-06, + "loss": 0.2262, + "num_input_tokens_seen": 9244584, + "step": 34470 + }, + { + "epoch": 9.036697247706423, + "grad_norm": 1.222638726234436, + "learning_rate": 1.4008398671645688e-06, + "loss": 0.145, + "num_input_tokens_seen": 9246040, + "step": 34475 + }, + { + "epoch": 9.038007863695936, + "grad_norm": 1.7763302326202393, + "learning_rate": 1.3970675585924241e-06, + "loss": 0.1216, + "num_input_tokens_seen": 9246936, + "step": 34480 + }, + { + "epoch": 9.039318479685452, + "grad_norm": 2.862910270690918, + "learning_rate": 1.3933001900732572e-06, + "loss": 0.2038, + "num_input_tokens_seen": 9248104, + "step": 34485 + }, + { + "epoch": 9.040629095674968, + "grad_norm": 1.6920785903930664, + "learning_rate": 1.3895377623955707e-06, + "loss": 0.1093, + "num_input_tokens_seen": 9249848, + "step": 34490 + }, + { + "epoch": 9.041939711664483, + "grad_norm": 1.7534067630767822, + "learning_rate": 1.3857802763468403e-06, + "loss": 0.1492, + "num_input_tokens_seen": 9251144, + "step": 34495 + }, + { + "epoch": 9.043250327653997, + "grad_norm": 2.3243932723999023, + "learning_rate": 1.3820277327134866e-06, + "loss": 0.1774, + "num_input_tokens_seen": 9252472, + "step": 34500 + }, + { + "epoch": 9.044560943643512, + "grad_norm": 1.7647031545639038, + "learning_rate": 1.3782801322809262e-06, + "loss": 0.1038, + "num_input_tokens_seen": 9253576, + "step": 34505 + }, + { + "epoch": 9.045871559633028, + "grad_norm": 3.5882887840270996, + "learning_rate": 1.374537475833515e-06, + "loss": 0.0978, + "num_input_tokens_seen": 9254968, + "step": 34510 + }, + { + "epoch": 9.047182175622543, + "grad_norm": 2.7857589721679688, + "learning_rate": 1.37079976415459e-06, + "loss": 0.3177, + "num_input_tokens_seen": 9256440, + "step": 34515 + }, + { + "epoch": 9.048492791612057, + "grad_norm": 1.4011293649673462, + "learning_rate": 1.3670669980264477e-06, + "loss": 0.102, + "num_input_tokens_seen": 9257688, + "step": 34520 + }, + { + "epoch": 9.049803407601573, + "grad_norm": 0.3536531329154968, + "learning_rate": 1.3633391782303468e-06, + "loss": 0.096, + "num_input_tokens_seen": 9258664, + "step": 34525 + }, + { + "epoch": 9.051114023591088, + "grad_norm": 3.125828981399536, + "learning_rate": 1.3596163055465154e-06, + "loss": 0.1393, + "num_input_tokens_seen": 9260344, + "step": 34530 + }, + { + "epoch": 9.052424639580602, + "grad_norm": 2.956669807434082, + "learning_rate": 1.3558983807541476e-06, + "loss": 0.2531, + "num_input_tokens_seen": 9263144, + "step": 34535 + }, + { + "epoch": 9.053735255570118, + "grad_norm": 1.932232141494751, + "learning_rate": 1.3521854046313897e-06, + "loss": 0.1483, + "num_input_tokens_seen": 9264136, + "step": 34540 + }, + { + "epoch": 9.055045871559633, + "grad_norm": 1.6771559715270996, + "learning_rate": 1.3484773779553677e-06, + "loss": 0.2474, + "num_input_tokens_seen": 9266360, + "step": 34545 + }, + { + "epoch": 9.056356487549149, + "grad_norm": 1.8568893671035767, + "learning_rate": 1.3447743015021636e-06, + "loss": 0.1734, + "num_input_tokens_seen": 9267624, + "step": 34550 + }, + { + "epoch": 9.057667103538662, + "grad_norm": 4.025647163391113, + "learning_rate": 1.3410761760468265e-06, + "loss": 0.2555, + "num_input_tokens_seen": 9269112, + "step": 34555 + }, + { + "epoch": 9.058977719528178, + "grad_norm": 1.2809675931930542, + "learning_rate": 1.3373830023633598e-06, + "loss": 0.1492, + "num_input_tokens_seen": 9270440, + "step": 34560 + }, + { + "epoch": 9.060288335517694, + "grad_norm": 1.9790037870407104, + "learning_rate": 1.3336947812247507e-06, + "loss": 0.0656, + "num_input_tokens_seen": 9271704, + "step": 34565 + }, + { + "epoch": 9.061598951507209, + "grad_norm": 2.123012065887451, + "learning_rate": 1.3300115134029345e-06, + "loss": 0.1005, + "num_input_tokens_seen": 9272984, + "step": 34570 + }, + { + "epoch": 9.062909567496723, + "grad_norm": 2.8834595680236816, + "learning_rate": 1.3263331996688055e-06, + "loss": 0.0792, + "num_input_tokens_seen": 9274488, + "step": 34575 + }, + { + "epoch": 9.064220183486238, + "grad_norm": 1.7015851736068726, + "learning_rate": 1.3226598407922342e-06, + "loss": 0.1451, + "num_input_tokens_seen": 9276504, + "step": 34580 + }, + { + "epoch": 9.065530799475754, + "grad_norm": 2.040672779083252, + "learning_rate": 1.318991437542047e-06, + "loss": 0.1227, + "num_input_tokens_seen": 9278200, + "step": 34585 + }, + { + "epoch": 9.06684141546527, + "grad_norm": 0.043829724192619324, + "learning_rate": 1.3153279906860355e-06, + "loss": 0.1853, + "num_input_tokens_seen": 9279880, + "step": 34590 + }, + { + "epoch": 9.068152031454783, + "grad_norm": 2.381303071975708, + "learning_rate": 1.3116695009909474e-06, + "loss": 0.0736, + "num_input_tokens_seen": 9280952, + "step": 34595 + }, + { + "epoch": 9.069462647444299, + "grad_norm": 2.260634422302246, + "learning_rate": 1.3080159692225063e-06, + "loss": 0.1508, + "num_input_tokens_seen": 9282008, + "step": 34600 + }, + { + "epoch": 9.070773263433814, + "grad_norm": 2.7745838165283203, + "learning_rate": 1.304367396145384e-06, + "loss": 0.1743, + "num_input_tokens_seen": 9283304, + "step": 34605 + }, + { + "epoch": 9.07208387942333, + "grad_norm": 4.329290866851807, + "learning_rate": 1.300723782523225e-06, + "loss": 0.1464, + "num_input_tokens_seen": 9284696, + "step": 34610 + }, + { + "epoch": 9.073394495412844, + "grad_norm": 3.4735934734344482, + "learning_rate": 1.2970851291186276e-06, + "loss": 0.1031, + "num_input_tokens_seen": 9285880, + "step": 34615 + }, + { + "epoch": 9.07470511140236, + "grad_norm": 2.2094149589538574, + "learning_rate": 1.2934514366931578e-06, + "loss": 0.0628, + "num_input_tokens_seen": 9287112, + "step": 34620 + }, + { + "epoch": 9.076015727391875, + "grad_norm": 1.25597083568573, + "learning_rate": 1.2898227060073403e-06, + "loss": 0.1496, + "num_input_tokens_seen": 9288488, + "step": 34625 + }, + { + "epoch": 9.077326343381388, + "grad_norm": 1.7777929306030273, + "learning_rate": 1.2861989378206624e-06, + "loss": 0.167, + "num_input_tokens_seen": 9290072, + "step": 34630 + }, + { + "epoch": 9.078636959370904, + "grad_norm": 1.6660560369491577, + "learning_rate": 1.2825801328915699e-06, + "loss": 0.1147, + "num_input_tokens_seen": 9291160, + "step": 34635 + }, + { + "epoch": 9.07994757536042, + "grad_norm": 2.9625144004821777, + "learning_rate": 1.2789662919774792e-06, + "loss": 0.1305, + "num_input_tokens_seen": 9292408, + "step": 34640 + }, + { + "epoch": 9.081258191349935, + "grad_norm": 3.7566144466400146, + "learning_rate": 1.2753574158347547e-06, + "loss": 0.2224, + "num_input_tokens_seen": 9294296, + "step": 34645 + }, + { + "epoch": 9.082568807339449, + "grad_norm": 2.6657283306121826, + "learning_rate": 1.271753505218734e-06, + "loss": 0.1947, + "num_input_tokens_seen": 9295256, + "step": 34650 + }, + { + "epoch": 9.083879423328964, + "grad_norm": 2.3655662536621094, + "learning_rate": 1.2681545608837026e-06, + "loss": 0.1681, + "num_input_tokens_seen": 9296696, + "step": 34655 + }, + { + "epoch": 9.08519003931848, + "grad_norm": 0.9890435934066772, + "learning_rate": 1.2645605835829132e-06, + "loss": 0.1845, + "num_input_tokens_seen": 9298792, + "step": 34660 + }, + { + "epoch": 9.086500655307995, + "grad_norm": 2.40847110748291, + "learning_rate": 1.2609715740685869e-06, + "loss": 0.1584, + "num_input_tokens_seen": 9300456, + "step": 34665 + }, + { + "epoch": 9.08781127129751, + "grad_norm": 1.705941915512085, + "learning_rate": 1.257387533091889e-06, + "loss": 0.1368, + "num_input_tokens_seen": 9302104, + "step": 34670 + }, + { + "epoch": 9.089121887287025, + "grad_norm": 0.6785320043563843, + "learning_rate": 1.2538084614029527e-06, + "loss": 0.1281, + "num_input_tokens_seen": 9303256, + "step": 34675 + }, + { + "epoch": 9.09043250327654, + "grad_norm": 2.285069227218628, + "learning_rate": 1.2502343597508792e-06, + "loss": 0.1144, + "num_input_tokens_seen": 9304776, + "step": 34680 + }, + { + "epoch": 9.091743119266056, + "grad_norm": 1.8162071704864502, + "learning_rate": 1.2466652288837227e-06, + "loss": 0.4498, + "num_input_tokens_seen": 9306296, + "step": 34685 + }, + { + "epoch": 9.09305373525557, + "grad_norm": 1.55417001247406, + "learning_rate": 1.2431010695484858e-06, + "loss": 0.2163, + "num_input_tokens_seen": 9307800, + "step": 34690 + }, + { + "epoch": 9.094364351245085, + "grad_norm": 0.9523571729660034, + "learning_rate": 1.2395418824911464e-06, + "loss": 0.1519, + "num_input_tokens_seen": 9309128, + "step": 34695 + }, + { + "epoch": 9.0956749672346, + "grad_norm": 1.956192135810852, + "learning_rate": 1.2359876684566368e-06, + "loss": 0.1188, + "num_input_tokens_seen": 9310520, + "step": 34700 + }, + { + "epoch": 9.096985583224116, + "grad_norm": 1.8630508184432983, + "learning_rate": 1.232438428188848e-06, + "loss": 0.1711, + "num_input_tokens_seen": 9311640, + "step": 34705 + }, + { + "epoch": 9.09829619921363, + "grad_norm": 6.474515914916992, + "learning_rate": 1.228894162430627e-06, + "loss": 0.0791, + "num_input_tokens_seen": 9312568, + "step": 34710 + }, + { + "epoch": 9.099606815203146, + "grad_norm": 1.3228486776351929, + "learning_rate": 1.225354871923784e-06, + "loss": 0.163, + "num_input_tokens_seen": 9314424, + "step": 34715 + }, + { + "epoch": 9.100917431192661, + "grad_norm": 1.426109790802002, + "learning_rate": 1.2218205574090896e-06, + "loss": 0.1428, + "num_input_tokens_seen": 9316008, + "step": 34720 + }, + { + "epoch": 9.102228047182175, + "grad_norm": 1.7403788566589355, + "learning_rate": 1.2182912196262664e-06, + "loss": 0.0967, + "num_input_tokens_seen": 9317128, + "step": 34725 + }, + { + "epoch": 9.10353866317169, + "grad_norm": 1.104697585105896, + "learning_rate": 1.2147668593139982e-06, + "loss": 0.1768, + "num_input_tokens_seen": 9318216, + "step": 34730 + }, + { + "epoch": 9.104849279161206, + "grad_norm": 4.1291728019714355, + "learning_rate": 1.2112474772099285e-06, + "loss": 0.1157, + "num_input_tokens_seen": 9319208, + "step": 34735 + }, + { + "epoch": 9.106159895150721, + "grad_norm": 2.329061269760132, + "learning_rate": 1.2077330740506598e-06, + "loss": 0.0961, + "num_input_tokens_seen": 9320248, + "step": 34740 + }, + { + "epoch": 9.107470511140235, + "grad_norm": 4.578170299530029, + "learning_rate": 1.2042236505717452e-06, + "loss": 0.1397, + "num_input_tokens_seen": 9321912, + "step": 34745 + }, + { + "epoch": 9.10878112712975, + "grad_norm": 2.85168194770813, + "learning_rate": 1.2007192075077085e-06, + "loss": 0.0796, + "num_input_tokens_seen": 9323080, + "step": 34750 + }, + { + "epoch": 9.110091743119266, + "grad_norm": 2.5300421714782715, + "learning_rate": 1.1972197455920181e-06, + "loss": 0.1885, + "num_input_tokens_seen": 9324328, + "step": 34755 + }, + { + "epoch": 9.111402359108782, + "grad_norm": 1.2259336709976196, + "learning_rate": 1.193725265557108e-06, + "loss": 0.0718, + "num_input_tokens_seen": 9325320, + "step": 34760 + }, + { + "epoch": 9.112712975098296, + "grad_norm": 2.9802684783935547, + "learning_rate": 1.1902357681343623e-06, + "loss": 0.1501, + "num_input_tokens_seen": 9326632, + "step": 34765 + }, + { + "epoch": 9.114023591087811, + "grad_norm": 2.7516186237335205, + "learning_rate": 1.1867512540541388e-06, + "loss": 0.1596, + "num_input_tokens_seen": 9327768, + "step": 34770 + }, + { + "epoch": 9.115334207077327, + "grad_norm": 1.8073866367340088, + "learning_rate": 1.1832717240457231e-06, + "loss": 0.1639, + "num_input_tokens_seen": 9329176, + "step": 34775 + }, + { + "epoch": 9.116644823066842, + "grad_norm": 2.1262121200561523, + "learning_rate": 1.1797971788373857e-06, + "loss": 0.117, + "num_input_tokens_seen": 9330504, + "step": 34780 + }, + { + "epoch": 9.117955439056356, + "grad_norm": 1.8344076871871948, + "learning_rate": 1.1763276191563422e-06, + "loss": 0.1127, + "num_input_tokens_seen": 9331768, + "step": 34785 + }, + { + "epoch": 9.119266055045872, + "grad_norm": 3.773528814315796, + "learning_rate": 1.1728630457287587e-06, + "loss": 0.1515, + "num_input_tokens_seen": 9332824, + "step": 34790 + }, + { + "epoch": 9.120576671035387, + "grad_norm": 2.6138668060302734, + "learning_rate": 1.1694034592797747e-06, + "loss": 0.1921, + "num_input_tokens_seen": 9334616, + "step": 34795 + }, + { + "epoch": 9.1218872870249, + "grad_norm": 2.058847427368164, + "learning_rate": 1.1659488605334695e-06, + "loss": 0.1543, + "num_input_tokens_seen": 9335864, + "step": 34800 + }, + { + "epoch": 9.123197903014416, + "grad_norm": 5.252617835998535, + "learning_rate": 1.1624992502128868e-06, + "loss": 0.2221, + "num_input_tokens_seen": 9337336, + "step": 34805 + }, + { + "epoch": 9.124508519003932, + "grad_norm": 1.0456889867782593, + "learning_rate": 1.159054629040024e-06, + "loss": 0.0952, + "num_input_tokens_seen": 9338600, + "step": 34810 + }, + { + "epoch": 9.125819134993447, + "grad_norm": 2.058854103088379, + "learning_rate": 1.1556149977358295e-06, + "loss": 0.0829, + "num_input_tokens_seen": 9339704, + "step": 34815 + }, + { + "epoch": 9.127129750982961, + "grad_norm": 0.564541757106781, + "learning_rate": 1.1521803570202188e-06, + "loss": 0.0637, + "num_input_tokens_seen": 9340712, + "step": 34820 + }, + { + "epoch": 9.128440366972477, + "grad_norm": 2.4368364810943604, + "learning_rate": 1.1487507076120507e-06, + "loss": 0.0985, + "num_input_tokens_seen": 9341640, + "step": 34825 + }, + { + "epoch": 9.129750982961992, + "grad_norm": 2.3895716667175293, + "learning_rate": 1.1453260502291507e-06, + "loss": 0.2647, + "num_input_tokens_seen": 9343096, + "step": 34830 + }, + { + "epoch": 9.131061598951508, + "grad_norm": 1.8227742910385132, + "learning_rate": 1.14190638558829e-06, + "loss": 0.1834, + "num_input_tokens_seen": 9344184, + "step": 34835 + }, + { + "epoch": 9.132372214941022, + "grad_norm": 1.2586380243301392, + "learning_rate": 1.1384917144051987e-06, + "loss": 0.1414, + "num_input_tokens_seen": 9345512, + "step": 34840 + }, + { + "epoch": 9.133682830930537, + "grad_norm": 1.0796332359313965, + "learning_rate": 1.135082037394561e-06, + "loss": 0.1609, + "num_input_tokens_seen": 9346888, + "step": 34845 + }, + { + "epoch": 9.134993446920053, + "grad_norm": 4.808542728424072, + "learning_rate": 1.1316773552700193e-06, + "loss": 0.1729, + "num_input_tokens_seen": 9348104, + "step": 34850 + }, + { + "epoch": 9.136304062909568, + "grad_norm": 2.2922704219818115, + "learning_rate": 1.1282776687441626e-06, + "loss": 0.2714, + "num_input_tokens_seen": 9349640, + "step": 34855 + }, + { + "epoch": 9.137614678899082, + "grad_norm": 1.2294855117797852, + "learning_rate": 1.1248829785285463e-06, + "loss": 0.1335, + "num_input_tokens_seen": 9350984, + "step": 34860 + }, + { + "epoch": 9.138925294888598, + "grad_norm": 1.7095363140106201, + "learning_rate": 1.121493285333669e-06, + "loss": 0.1795, + "num_input_tokens_seen": 9352200, + "step": 34865 + }, + { + "epoch": 9.140235910878113, + "grad_norm": 2.7844042778015137, + "learning_rate": 1.1181085898689881e-06, + "loss": 0.1716, + "num_input_tokens_seen": 9353656, + "step": 34870 + }, + { + "epoch": 9.141546526867629, + "grad_norm": 0.694219708442688, + "learning_rate": 1.1147288928429118e-06, + "loss": 0.1022, + "num_input_tokens_seen": 9355032, + "step": 34875 + }, + { + "epoch": 9.142857142857142, + "grad_norm": 1.6717454195022583, + "learning_rate": 1.1113541949628104e-06, + "loss": 0.1428, + "num_input_tokens_seen": 9356248, + "step": 34880 + }, + { + "epoch": 9.144167758846658, + "grad_norm": 1.847715973854065, + "learning_rate": 1.1079844969350023e-06, + "loss": 0.0908, + "num_input_tokens_seen": 9357480, + "step": 34885 + }, + { + "epoch": 9.145478374836173, + "grad_norm": 3.011017322540283, + "learning_rate": 1.104619799464754e-06, + "loss": 0.0967, + "num_input_tokens_seen": 9358472, + "step": 34890 + }, + { + "epoch": 9.146788990825687, + "grad_norm": 4.326774597167969, + "learning_rate": 1.1012601032562935e-06, + "loss": 0.1963, + "num_input_tokens_seen": 9360072, + "step": 34895 + }, + { + "epoch": 9.148099606815203, + "grad_norm": 0.9527531862258911, + "learning_rate": 1.0979054090127978e-06, + "loss": 0.2048, + "num_input_tokens_seen": 9361432, + "step": 34900 + }, + { + "epoch": 9.149410222804718, + "grad_norm": 1.089834213256836, + "learning_rate": 1.094555717436399e-06, + "loss": 0.1228, + "num_input_tokens_seen": 9362600, + "step": 34905 + }, + { + "epoch": 9.150720838794234, + "grad_norm": 1.6916688680648804, + "learning_rate": 1.0912110292281868e-06, + "loss": 0.1344, + "num_input_tokens_seen": 9363832, + "step": 34910 + }, + { + "epoch": 9.152031454783748, + "grad_norm": 2.1484665870666504, + "learning_rate": 1.0878713450881928e-06, + "loss": 0.1501, + "num_input_tokens_seen": 9365160, + "step": 34915 + }, + { + "epoch": 9.153342070773263, + "grad_norm": 2.712648630142212, + "learning_rate": 1.084536665715416e-06, + "loss": 0.538, + "num_input_tokens_seen": 9366504, + "step": 34920 + }, + { + "epoch": 9.154652686762779, + "grad_norm": 1.7845371961593628, + "learning_rate": 1.0812069918077844e-06, + "loss": 0.1288, + "num_input_tokens_seen": 9367768, + "step": 34925 + }, + { + "epoch": 9.155963302752294, + "grad_norm": 3.0990734100341797, + "learning_rate": 1.0778823240622043e-06, + "loss": 0.0966, + "num_input_tokens_seen": 9368872, + "step": 34930 + }, + { + "epoch": 9.157273918741808, + "grad_norm": 2.854763984680176, + "learning_rate": 1.0745626631745193e-06, + "loss": 0.2438, + "num_input_tokens_seen": 9371912, + "step": 34935 + }, + { + "epoch": 9.158584534731324, + "grad_norm": 2.2229666709899902, + "learning_rate": 1.0712480098395289e-06, + "loss": 0.1814, + "num_input_tokens_seen": 9373624, + "step": 34940 + }, + { + "epoch": 9.159895150720839, + "grad_norm": 0.6518092751502991, + "learning_rate": 1.0679383647509839e-06, + "loss": 0.0823, + "num_input_tokens_seen": 9374984, + "step": 34945 + }, + { + "epoch": 9.161205766710355, + "grad_norm": 2.3854711055755615, + "learning_rate": 1.0646337286015856e-06, + "loss": 0.1105, + "num_input_tokens_seen": 9375880, + "step": 34950 + }, + { + "epoch": 9.162516382699868, + "grad_norm": 1.7810224294662476, + "learning_rate": 1.0613341020829948e-06, + "loss": 0.1656, + "num_input_tokens_seen": 9377112, + "step": 34955 + }, + { + "epoch": 9.163826998689384, + "grad_norm": 1.8214871883392334, + "learning_rate": 1.0580394858858117e-06, + "loss": 0.1097, + "num_input_tokens_seen": 9378120, + "step": 34960 + }, + { + "epoch": 9.1651376146789, + "grad_norm": 1.1339293718338013, + "learning_rate": 1.054749880699593e-06, + "loss": 0.0706, + "num_input_tokens_seen": 9379192, + "step": 34965 + }, + { + "epoch": 9.166448230668415, + "grad_norm": 1.8644287586212158, + "learning_rate": 1.051465287212855e-06, + "loss": 0.1327, + "num_input_tokens_seen": 9380312, + "step": 34970 + }, + { + "epoch": 9.167758846657929, + "grad_norm": 0.6601839661598206, + "learning_rate": 1.04818570611305e-06, + "loss": 0.1497, + "num_input_tokens_seen": 9381976, + "step": 34975 + }, + { + "epoch": 9.169069462647444, + "grad_norm": 1.6255574226379395, + "learning_rate": 1.0449111380865906e-06, + "loss": 0.0621, + "num_input_tokens_seen": 9383000, + "step": 34980 + }, + { + "epoch": 9.17038007863696, + "grad_norm": 1.8832076787948608, + "learning_rate": 1.0416415838188393e-06, + "loss": 0.2581, + "num_input_tokens_seen": 9384264, + "step": 34985 + }, + { + "epoch": 9.171690694626474, + "grad_norm": 1.0596939325332642, + "learning_rate": 1.038377043994107e-06, + "loss": 0.0427, + "num_input_tokens_seen": 9385640, + "step": 34990 + }, + { + "epoch": 9.17300131061599, + "grad_norm": 2.168516159057617, + "learning_rate": 1.0351175192956585e-06, + "loss": 0.1769, + "num_input_tokens_seen": 9387160, + "step": 34995 + }, + { + "epoch": 9.174311926605505, + "grad_norm": 4.104536056518555, + "learning_rate": 1.0318630104057037e-06, + "loss": 0.2095, + "num_input_tokens_seen": 9388248, + "step": 35000 + }, + { + "epoch": 9.17562254259502, + "grad_norm": 0.21695250272750854, + "learning_rate": 1.0286135180054114e-06, + "loss": 0.0839, + "num_input_tokens_seen": 9389480, + "step": 35005 + }, + { + "epoch": 9.176933158584534, + "grad_norm": 3.4998600482940674, + "learning_rate": 1.0253690427748874e-06, + "loss": 0.2077, + "num_input_tokens_seen": 9390856, + "step": 35010 + }, + { + "epoch": 9.17824377457405, + "grad_norm": 2.7096428871154785, + "learning_rate": 1.0221295853931973e-06, + "loss": 0.0754, + "num_input_tokens_seen": 9392488, + "step": 35015 + }, + { + "epoch": 9.179554390563565, + "grad_norm": 1.8937492370605469, + "learning_rate": 1.0188951465383511e-06, + "loss": 0.1409, + "num_input_tokens_seen": 9394072, + "step": 35020 + }, + { + "epoch": 9.18086500655308, + "grad_norm": 1.2566646337509155, + "learning_rate": 1.015665726887316e-06, + "loss": 0.08, + "num_input_tokens_seen": 9395288, + "step": 35025 + }, + { + "epoch": 9.182175622542594, + "grad_norm": 3.772820234298706, + "learning_rate": 1.0124413271160038e-06, + "loss": 0.1743, + "num_input_tokens_seen": 9396472, + "step": 35030 + }, + { + "epoch": 9.18348623853211, + "grad_norm": 3.2302026748657227, + "learning_rate": 1.0092219478992776e-06, + "loss": 0.1195, + "num_input_tokens_seen": 9397832, + "step": 35035 + }, + { + "epoch": 9.184796854521625, + "grad_norm": 2.15102219581604, + "learning_rate": 1.0060075899109427e-06, + "loss": 0.1758, + "num_input_tokens_seen": 9399256, + "step": 35040 + }, + { + "epoch": 9.186107470511141, + "grad_norm": 1.1934549808502197, + "learning_rate": 1.0027982538237612e-06, + "loss": 0.1686, + "num_input_tokens_seen": 9401176, + "step": 35045 + }, + { + "epoch": 9.187418086500655, + "grad_norm": 3.2430245876312256, + "learning_rate": 9.995939403094402e-07, + "loss": 0.1131, + "num_input_tokens_seen": 9402280, + "step": 35050 + }, + { + "epoch": 9.18872870249017, + "grad_norm": 1.9667595624923706, + "learning_rate": 9.963946500386374e-07, + "loss": 0.1472, + "num_input_tokens_seen": 9403352, + "step": 35055 + }, + { + "epoch": 9.190039318479686, + "grad_norm": 2.923372983932495, + "learning_rate": 9.932003836809616e-07, + "loss": 0.2305, + "num_input_tokens_seen": 9404568, + "step": 35060 + }, + { + "epoch": 9.191349934469201, + "grad_norm": 5.209006309509277, + "learning_rate": 9.900111419049619e-07, + "loss": 0.0603, + "num_input_tokens_seen": 9405496, + "step": 35065 + }, + { + "epoch": 9.192660550458715, + "grad_norm": 2.4664108753204346, + "learning_rate": 9.868269253781453e-07, + "loss": 0.126, + "num_input_tokens_seen": 9406904, + "step": 35070 + }, + { + "epoch": 9.19397116644823, + "grad_norm": 2.997520685195923, + "learning_rate": 9.836477347669626e-07, + "loss": 0.1586, + "num_input_tokens_seen": 9408200, + "step": 35075 + }, + { + "epoch": 9.195281782437746, + "grad_norm": 2.0632293224334717, + "learning_rate": 9.804735707368118e-07, + "loss": 0.1456, + "num_input_tokens_seen": 9409448, + "step": 35080 + }, + { + "epoch": 9.19659239842726, + "grad_norm": 2.4153800010681152, + "learning_rate": 9.773044339520392e-07, + "loss": 0.1508, + "num_input_tokens_seen": 9410888, + "step": 35085 + }, + { + "epoch": 9.197903014416775, + "grad_norm": 2.514216661453247, + "learning_rate": 9.741403250759424e-07, + "loss": 0.1031, + "num_input_tokens_seen": 9412008, + "step": 35090 + }, + { + "epoch": 9.199213630406291, + "grad_norm": 2.320625066757202, + "learning_rate": 9.709812447707606e-07, + "loss": 0.1828, + "num_input_tokens_seen": 9413304, + "step": 35095 + }, + { + "epoch": 9.200524246395807, + "grad_norm": 3.366960287094116, + "learning_rate": 9.678271936976847e-07, + "loss": 0.1257, + "num_input_tokens_seen": 9414440, + "step": 35100 + }, + { + "epoch": 9.20183486238532, + "grad_norm": 2.338993787765503, + "learning_rate": 9.64678172516853e-07, + "loss": 0.1304, + "num_input_tokens_seen": 9415672, + "step": 35105 + }, + { + "epoch": 9.203145478374836, + "grad_norm": 1.918689250946045, + "learning_rate": 9.615341818873496e-07, + "loss": 0.0989, + "num_input_tokens_seen": 9416872, + "step": 35110 + }, + { + "epoch": 9.204456094364351, + "grad_norm": 4.246461868286133, + "learning_rate": 9.583952224672094e-07, + "loss": 0.1792, + "num_input_tokens_seen": 9417976, + "step": 35115 + }, + { + "epoch": 9.205766710353867, + "grad_norm": 1.4985146522521973, + "learning_rate": 9.552612949134094e-07, + "loss": 0.1556, + "num_input_tokens_seen": 9419496, + "step": 35120 + }, + { + "epoch": 9.20707732634338, + "grad_norm": 1.6359652280807495, + "learning_rate": 9.521323998818694e-07, + "loss": 0.1257, + "num_input_tokens_seen": 9420536, + "step": 35125 + }, + { + "epoch": 9.208387942332896, + "grad_norm": 2.626540422439575, + "learning_rate": 9.490085380274682e-07, + "loss": 0.1689, + "num_input_tokens_seen": 9422248, + "step": 35130 + }, + { + "epoch": 9.209698558322412, + "grad_norm": 0.7842778563499451, + "learning_rate": 9.45889710004022e-07, + "loss": 0.1055, + "num_input_tokens_seen": 9423176, + "step": 35135 + }, + { + "epoch": 9.211009174311927, + "grad_norm": 2.054986000061035, + "learning_rate": 9.427759164642974e-07, + "loss": 0.149, + "num_input_tokens_seen": 9424360, + "step": 35140 + }, + { + "epoch": 9.212319790301441, + "grad_norm": 1.194831371307373, + "learning_rate": 9.396671580600064e-07, + "loss": 0.0628, + "num_input_tokens_seen": 9425384, + "step": 35145 + }, + { + "epoch": 9.213630406290957, + "grad_norm": 0.95365971326828, + "learning_rate": 9.36563435441809e-07, + "loss": 0.1047, + "num_input_tokens_seen": 9427304, + "step": 35150 + }, + { + "epoch": 9.214941022280472, + "grad_norm": 2.0964677333831787, + "learning_rate": 9.334647492593051e-07, + "loss": 0.1065, + "num_input_tokens_seen": 9428472, + "step": 35155 + }, + { + "epoch": 9.216251638269988, + "grad_norm": 2.6577160358428955, + "learning_rate": 9.303711001610454e-07, + "loss": 0.2303, + "num_input_tokens_seen": 9429576, + "step": 35160 + }, + { + "epoch": 9.217562254259501, + "grad_norm": 0.2935457229614258, + "learning_rate": 9.27282488794523e-07, + "loss": 0.0546, + "num_input_tokens_seen": 9430504, + "step": 35165 + }, + { + "epoch": 9.218872870249017, + "grad_norm": 1.879996657371521, + "learning_rate": 9.24198915806182e-07, + "loss": 0.1498, + "num_input_tokens_seen": 9432408, + "step": 35170 + }, + { + "epoch": 9.220183486238533, + "grad_norm": 1.6692171096801758, + "learning_rate": 9.211203818414088e-07, + "loss": 0.1078, + "num_input_tokens_seen": 9433592, + "step": 35175 + }, + { + "epoch": 9.221494102228046, + "grad_norm": 2.5598037242889404, + "learning_rate": 9.180468875445352e-07, + "loss": 0.1081, + "num_input_tokens_seen": 9434648, + "step": 35180 + }, + { + "epoch": 9.222804718217562, + "grad_norm": 1.899776816368103, + "learning_rate": 9.149784335588357e-07, + "loss": 0.1067, + "num_input_tokens_seen": 9436152, + "step": 35185 + }, + { + "epoch": 9.224115334207077, + "grad_norm": 4.596175670623779, + "learning_rate": 9.119150205265326e-07, + "loss": 0.0968, + "num_input_tokens_seen": 9437336, + "step": 35190 + }, + { + "epoch": 9.225425950196593, + "grad_norm": 1.3031920194625854, + "learning_rate": 9.088566490888017e-07, + "loss": 0.1056, + "num_input_tokens_seen": 9438728, + "step": 35195 + }, + { + "epoch": 9.226736566186107, + "grad_norm": 1.384186029434204, + "learning_rate": 9.058033198857424e-07, + "loss": 0.1774, + "num_input_tokens_seen": 9440216, + "step": 35200 + }, + { + "epoch": 9.228047182175622, + "grad_norm": 4.106912612915039, + "learning_rate": 9.027550335564184e-07, + "loss": 0.1298, + "num_input_tokens_seen": 9441208, + "step": 35205 + }, + { + "epoch": 9.229357798165138, + "grad_norm": 1.8522605895996094, + "learning_rate": 8.997117907388275e-07, + "loss": 0.1399, + "num_input_tokens_seen": 9442152, + "step": 35210 + }, + { + "epoch": 9.230668414154653, + "grad_norm": 0.515232503414154, + "learning_rate": 8.966735920699187e-07, + "loss": 0.1252, + "num_input_tokens_seen": 9443528, + "step": 35215 + }, + { + "epoch": 9.231979030144167, + "grad_norm": 0.1569172888994217, + "learning_rate": 8.936404381855806e-07, + "loss": 0.0955, + "num_input_tokens_seen": 9444488, + "step": 35220 + }, + { + "epoch": 9.233289646133683, + "grad_norm": 2.5804104804992676, + "learning_rate": 8.906123297206442e-07, + "loss": 0.1082, + "num_input_tokens_seen": 9445720, + "step": 35225 + }, + { + "epoch": 9.234600262123198, + "grad_norm": 1.0356110334396362, + "learning_rate": 8.875892673088887e-07, + "loss": 0.13, + "num_input_tokens_seen": 9447064, + "step": 35230 + }, + { + "epoch": 9.235910878112714, + "grad_norm": 2.67598032951355, + "learning_rate": 8.845712515830385e-07, + "loss": 0.1754, + "num_input_tokens_seen": 9448392, + "step": 35235 + }, + { + "epoch": 9.237221494102227, + "grad_norm": 2.494436264038086, + "learning_rate": 8.815582831747576e-07, + "loss": 0.1675, + "num_input_tokens_seen": 9449912, + "step": 35240 + }, + { + "epoch": 9.238532110091743, + "grad_norm": 0.6476318836212158, + "learning_rate": 8.785503627146501e-07, + "loss": 0.2947, + "num_input_tokens_seen": 9451560, + "step": 35245 + }, + { + "epoch": 9.239842726081259, + "grad_norm": 1.3266923427581787, + "learning_rate": 8.755474908322764e-07, + "loss": 0.1788, + "num_input_tokens_seen": 9452984, + "step": 35250 + }, + { + "epoch": 9.241153342070774, + "grad_norm": 2.6574175357818604, + "learning_rate": 8.725496681561279e-07, + "loss": 0.2581, + "num_input_tokens_seen": 9454568, + "step": 35255 + }, + { + "epoch": 9.242463958060288, + "grad_norm": 3.6212100982666016, + "learning_rate": 8.695568953136446e-07, + "loss": 0.1898, + "num_input_tokens_seen": 9455880, + "step": 35260 + }, + { + "epoch": 9.243774574049803, + "grad_norm": 2.914926767349243, + "learning_rate": 8.665691729312115e-07, + "loss": 0.1789, + "num_input_tokens_seen": 9457624, + "step": 35265 + }, + { + "epoch": 9.245085190039319, + "grad_norm": 1.5759066343307495, + "learning_rate": 8.635865016341477e-07, + "loss": 0.092, + "num_input_tokens_seen": 9459352, + "step": 35270 + }, + { + "epoch": 9.246395806028833, + "grad_norm": 2.4548394680023193, + "learning_rate": 8.606088820467318e-07, + "loss": 0.0666, + "num_input_tokens_seen": 9460328, + "step": 35275 + }, + { + "epoch": 9.247706422018348, + "grad_norm": 2.052332639694214, + "learning_rate": 8.576363147921596e-07, + "loss": 0.2873, + "num_input_tokens_seen": 9461880, + "step": 35280 + }, + { + "epoch": 9.249017038007864, + "grad_norm": 1.2548021078109741, + "learning_rate": 8.546688004925946e-07, + "loss": 0.0687, + "num_input_tokens_seen": 9463336, + "step": 35285 + }, + { + "epoch": 9.25032765399738, + "grad_norm": 2.0489790439605713, + "learning_rate": 8.517063397691288e-07, + "loss": 0.1179, + "num_input_tokens_seen": 9465400, + "step": 35290 + }, + { + "epoch": 9.251638269986893, + "grad_norm": 3.9518771171569824, + "learning_rate": 8.487489332418025e-07, + "loss": 0.1947, + "num_input_tokens_seen": 9467048, + "step": 35295 + }, + { + "epoch": 9.252948885976409, + "grad_norm": 1.8111047744750977, + "learning_rate": 8.457965815295926e-07, + "loss": 0.0981, + "num_input_tokens_seen": 9468104, + "step": 35300 + }, + { + "epoch": 9.254259501965924, + "grad_norm": 0.8033950328826904, + "learning_rate": 8.428492852504216e-07, + "loss": 0.165, + "num_input_tokens_seen": 9469128, + "step": 35305 + }, + { + "epoch": 9.25557011795544, + "grad_norm": 2.8473615646362305, + "learning_rate": 8.399070450211599e-07, + "loss": 0.1127, + "num_input_tokens_seen": 9470088, + "step": 35310 + }, + { + "epoch": 9.256880733944953, + "grad_norm": 1.642051339149475, + "learning_rate": 8.369698614576038e-07, + "loss": 0.1181, + "num_input_tokens_seen": 9471208, + "step": 35315 + }, + { + "epoch": 9.258191349934469, + "grad_norm": 1.704128384590149, + "learning_rate": 8.340377351745088e-07, + "loss": 0.3653, + "num_input_tokens_seen": 9472728, + "step": 35320 + }, + { + "epoch": 9.259501965923985, + "grad_norm": 0.04663309454917908, + "learning_rate": 8.311106667855562e-07, + "loss": 0.082, + "num_input_tokens_seen": 9474008, + "step": 35325 + }, + { + "epoch": 9.2608125819135, + "grad_norm": 1.4947171211242676, + "learning_rate": 8.281886569033836e-07, + "loss": 0.1496, + "num_input_tokens_seen": 9475528, + "step": 35330 + }, + { + "epoch": 9.262123197903014, + "grad_norm": 3.122575521469116, + "learning_rate": 8.252717061395576e-07, + "loss": 0.3915, + "num_input_tokens_seen": 9476856, + "step": 35335 + }, + { + "epoch": 9.26343381389253, + "grad_norm": 3.147019863128662, + "learning_rate": 8.223598151045952e-07, + "loss": 0.0898, + "num_input_tokens_seen": 9477880, + "step": 35340 + }, + { + "epoch": 9.264744429882045, + "grad_norm": 3.035698175430298, + "learning_rate": 8.194529844079451e-07, + "loss": 0.201, + "num_input_tokens_seen": 9479304, + "step": 35345 + }, + { + "epoch": 9.26605504587156, + "grad_norm": 1.566272497177124, + "learning_rate": 8.16551214658004e-07, + "loss": 0.1295, + "num_input_tokens_seen": 9480280, + "step": 35350 + }, + { + "epoch": 9.267365661861074, + "grad_norm": 0.3679353892803192, + "learning_rate": 8.136545064621137e-07, + "loss": 0.1772, + "num_input_tokens_seen": 9481496, + "step": 35355 + }, + { + "epoch": 9.26867627785059, + "grad_norm": 2.5467543601989746, + "learning_rate": 8.107628604265366e-07, + "loss": 0.1155, + "num_input_tokens_seen": 9483144, + "step": 35360 + }, + { + "epoch": 9.269986893840105, + "grad_norm": 1.7746244668960571, + "learning_rate": 8.078762771564996e-07, + "loss": 0.1251, + "num_input_tokens_seen": 9484616, + "step": 35365 + }, + { + "epoch": 9.271297509829619, + "grad_norm": 2.573075771331787, + "learning_rate": 8.049947572561584e-07, + "loss": 0.1106, + "num_input_tokens_seen": 9485800, + "step": 35370 + }, + { + "epoch": 9.272608125819135, + "grad_norm": 1.0471751689910889, + "learning_rate": 8.021183013286082e-07, + "loss": 0.1116, + "num_input_tokens_seen": 9486984, + "step": 35375 + }, + { + "epoch": 9.27391874180865, + "grad_norm": 2.3810551166534424, + "learning_rate": 7.99246909975887e-07, + "loss": 0.2669, + "num_input_tokens_seen": 9488632, + "step": 35380 + }, + { + "epoch": 9.275229357798166, + "grad_norm": 1.5829968452453613, + "learning_rate": 7.963805837989724e-07, + "loss": 0.2114, + "num_input_tokens_seen": 9489880, + "step": 35385 + }, + { + "epoch": 9.27653997378768, + "grad_norm": 1.578926682472229, + "learning_rate": 7.935193233977845e-07, + "loss": 0.1272, + "num_input_tokens_seen": 9491032, + "step": 35390 + }, + { + "epoch": 9.277850589777195, + "grad_norm": 2.6898951530456543, + "learning_rate": 7.906631293711719e-07, + "loss": 0.1159, + "num_input_tokens_seen": 9492520, + "step": 35395 + }, + { + "epoch": 9.27916120576671, + "grad_norm": 5.012596607208252, + "learning_rate": 7.878120023169344e-07, + "loss": 0.1288, + "num_input_tokens_seen": 9494040, + "step": 35400 + }, + { + "epoch": 9.280471821756226, + "grad_norm": 1.8940261602401733, + "learning_rate": 7.849659428318113e-07, + "loss": 0.1083, + "num_input_tokens_seen": 9495048, + "step": 35405 + }, + { + "epoch": 9.28178243774574, + "grad_norm": 1.8796881437301636, + "learning_rate": 7.821249515114787e-07, + "loss": 0.175, + "num_input_tokens_seen": 9496776, + "step": 35410 + }, + { + "epoch": 9.283093053735255, + "grad_norm": 2.1169354915618896, + "learning_rate": 7.792890289505444e-07, + "loss": 0.2621, + "num_input_tokens_seen": 9498040, + "step": 35415 + }, + { + "epoch": 9.284403669724771, + "grad_norm": 1.064894676208496, + "learning_rate": 7.764581757425642e-07, + "loss": 0.0796, + "num_input_tokens_seen": 9499608, + "step": 35420 + }, + { + "epoch": 9.285714285714286, + "grad_norm": 3.0066821575164795, + "learning_rate": 7.736323924800365e-07, + "loss": 0.2254, + "num_input_tokens_seen": 9500984, + "step": 35425 + }, + { + "epoch": 9.2870249017038, + "grad_norm": 1.5966204404830933, + "learning_rate": 7.708116797543907e-07, + "loss": 0.0969, + "num_input_tokens_seen": 9502184, + "step": 35430 + }, + { + "epoch": 9.288335517693316, + "grad_norm": 1.935320496559143, + "learning_rate": 7.679960381559936e-07, + "loss": 0.2242, + "num_input_tokens_seen": 9503624, + "step": 35435 + }, + { + "epoch": 9.289646133682831, + "grad_norm": 1.5240797996520996, + "learning_rate": 7.651854682741572e-07, + "loss": 0.1514, + "num_input_tokens_seen": 9505176, + "step": 35440 + }, + { + "epoch": 9.290956749672347, + "grad_norm": 1.0899872779846191, + "learning_rate": 7.623799706971274e-07, + "loss": 0.1911, + "num_input_tokens_seen": 9506504, + "step": 35445 + }, + { + "epoch": 9.29226736566186, + "grad_norm": 2.6870548725128174, + "learning_rate": 7.595795460120903e-07, + "loss": 0.1069, + "num_input_tokens_seen": 9507448, + "step": 35450 + }, + { + "epoch": 9.293577981651376, + "grad_norm": 1.8906943798065186, + "learning_rate": 7.567841948051685e-07, + "loss": 0.1609, + "num_input_tokens_seen": 9508760, + "step": 35455 + }, + { + "epoch": 9.294888597640892, + "grad_norm": 2.8955719470977783, + "learning_rate": 7.539939176614247e-07, + "loss": 0.1565, + "num_input_tokens_seen": 9510072, + "step": 35460 + }, + { + "epoch": 9.296199213630405, + "grad_norm": 0.9994878172874451, + "learning_rate": 7.512087151648639e-07, + "loss": 0.1081, + "num_input_tokens_seen": 9511288, + "step": 35465 + }, + { + "epoch": 9.297509829619921, + "grad_norm": 1.7661978006362915, + "learning_rate": 7.484285878984171e-07, + "loss": 0.2059, + "num_input_tokens_seen": 9512408, + "step": 35470 + }, + { + "epoch": 9.298820445609437, + "grad_norm": 2.058985948562622, + "learning_rate": 7.45653536443966e-07, + "loss": 0.1389, + "num_input_tokens_seen": 9513752, + "step": 35475 + }, + { + "epoch": 9.300131061598952, + "grad_norm": 2.5706048011779785, + "learning_rate": 7.428835613823182e-07, + "loss": 0.1961, + "num_input_tokens_seen": 9515000, + "step": 35480 + }, + { + "epoch": 9.301441677588466, + "grad_norm": 1.4546998739242554, + "learning_rate": 7.401186632932294e-07, + "loss": 0.1216, + "num_input_tokens_seen": 9516136, + "step": 35485 + }, + { + "epoch": 9.302752293577981, + "grad_norm": 1.935347557067871, + "learning_rate": 7.373588427553869e-07, + "loss": 0.1386, + "num_input_tokens_seen": 9517288, + "step": 35490 + }, + { + "epoch": 9.304062909567497, + "grad_norm": 2.3806324005126953, + "learning_rate": 7.346041003464172e-07, + "loss": 0.1347, + "num_input_tokens_seen": 9518536, + "step": 35495 + }, + { + "epoch": 9.305373525557012, + "grad_norm": 1.9547449350357056, + "learning_rate": 7.318544366428814e-07, + "loss": 0.1185, + "num_input_tokens_seen": 9519608, + "step": 35500 + }, + { + "epoch": 9.306684141546526, + "grad_norm": 2.2677221298217773, + "learning_rate": 7.291098522202777e-07, + "loss": 0.1363, + "num_input_tokens_seen": 9520808, + "step": 35505 + }, + { + "epoch": 9.307994757536042, + "grad_norm": 1.5686860084533691, + "learning_rate": 7.263703476530492e-07, + "loss": 0.1473, + "num_input_tokens_seen": 9522008, + "step": 35510 + }, + { + "epoch": 9.309305373525557, + "grad_norm": 1.8175143003463745, + "learning_rate": 7.236359235145624e-07, + "loss": 0.1772, + "num_input_tokens_seen": 9523288, + "step": 35515 + }, + { + "epoch": 9.310615989515073, + "grad_norm": 2.5036957263946533, + "learning_rate": 7.209065803771315e-07, + "loss": 0.1577, + "num_input_tokens_seen": 9525064, + "step": 35520 + }, + { + "epoch": 9.311926605504587, + "grad_norm": 3.756870985031128, + "learning_rate": 7.181823188120024e-07, + "loss": 0.18, + "num_input_tokens_seen": 9526120, + "step": 35525 + }, + { + "epoch": 9.313237221494102, + "grad_norm": 1.7306586503982544, + "learning_rate": 7.154631393893552e-07, + "loss": 0.1684, + "num_input_tokens_seen": 9527512, + "step": 35530 + }, + { + "epoch": 9.314547837483618, + "grad_norm": 1.268473505973816, + "learning_rate": 7.127490426783123e-07, + "loss": 0.2159, + "num_input_tokens_seen": 9528600, + "step": 35535 + }, + { + "epoch": 9.315858453473133, + "grad_norm": 1.832426905632019, + "learning_rate": 7.100400292469333e-07, + "loss": 0.2346, + "num_input_tokens_seen": 9529800, + "step": 35540 + }, + { + "epoch": 9.317169069462647, + "grad_norm": 1.8149654865264893, + "learning_rate": 7.073360996622064e-07, + "loss": 0.1045, + "num_input_tokens_seen": 9530968, + "step": 35545 + }, + { + "epoch": 9.318479685452163, + "grad_norm": 1.6654456853866577, + "learning_rate": 7.046372544900592e-07, + "loss": 0.0818, + "num_input_tokens_seen": 9532168, + "step": 35550 + }, + { + "epoch": 9.319790301441678, + "grad_norm": 1.3666173219680786, + "learning_rate": 7.01943494295354e-07, + "loss": 0.1851, + "num_input_tokens_seen": 9533640, + "step": 35555 + }, + { + "epoch": 9.321100917431192, + "grad_norm": 2.905989408493042, + "learning_rate": 6.992548196418924e-07, + "loss": 0.0855, + "num_input_tokens_seen": 9534632, + "step": 35560 + }, + { + "epoch": 9.322411533420707, + "grad_norm": 1.6682051420211792, + "learning_rate": 6.965712310924078e-07, + "loss": 0.1165, + "num_input_tokens_seen": 9536040, + "step": 35565 + }, + { + "epoch": 9.323722149410223, + "grad_norm": 1.8770228624343872, + "learning_rate": 6.9389272920857e-07, + "loss": 0.0897, + "num_input_tokens_seen": 9537192, + "step": 35570 + }, + { + "epoch": 9.325032765399738, + "grad_norm": 1.2881629467010498, + "learning_rate": 6.912193145509893e-07, + "loss": 0.0836, + "num_input_tokens_seen": 9538680, + "step": 35575 + }, + { + "epoch": 9.326343381389252, + "grad_norm": 1.5408869981765747, + "learning_rate": 6.885509876792012e-07, + "loss": 0.0952, + "num_input_tokens_seen": 9539656, + "step": 35580 + }, + { + "epoch": 9.327653997378768, + "grad_norm": 4.114366054534912, + "learning_rate": 6.858877491516869e-07, + "loss": 0.137, + "num_input_tokens_seen": 9540968, + "step": 35585 + }, + { + "epoch": 9.328964613368283, + "grad_norm": 2.2682487964630127, + "learning_rate": 6.832295995258531e-07, + "loss": 0.1904, + "num_input_tokens_seen": 9542248, + "step": 35590 + }, + { + "epoch": 9.330275229357799, + "grad_norm": 2.339280128479004, + "learning_rate": 6.805765393580493e-07, + "loss": 0.1583, + "num_input_tokens_seen": 9543416, + "step": 35595 + }, + { + "epoch": 9.331585845347313, + "grad_norm": 2.001277208328247, + "learning_rate": 6.779285692035536e-07, + "loss": 0.1955, + "num_input_tokens_seen": 9545272, + "step": 35600 + }, + { + "epoch": 9.332896461336828, + "grad_norm": 2.008765935897827, + "learning_rate": 6.752856896165866e-07, + "loss": 0.1872, + "num_input_tokens_seen": 9546488, + "step": 35605 + }, + { + "epoch": 9.334207077326344, + "grad_norm": 1.5906059741973877, + "learning_rate": 6.726479011502917e-07, + "loss": 0.0797, + "num_input_tokens_seen": 9547496, + "step": 35610 + }, + { + "epoch": 9.33551769331586, + "grad_norm": 3.277066707611084, + "learning_rate": 6.700152043567554e-07, + "loss": 0.1209, + "num_input_tokens_seen": 9548856, + "step": 35615 + }, + { + "epoch": 9.336828309305373, + "grad_norm": 3.2773401737213135, + "learning_rate": 6.673875997870005e-07, + "loss": 0.1153, + "num_input_tokens_seen": 9550280, + "step": 35620 + }, + { + "epoch": 9.338138925294889, + "grad_norm": 1.3432507514953613, + "learning_rate": 6.64765087990979e-07, + "loss": 0.2011, + "num_input_tokens_seen": 9551864, + "step": 35625 + }, + { + "epoch": 9.339449541284404, + "grad_norm": 1.5712320804595947, + "learning_rate": 6.621476695175738e-07, + "loss": 0.1104, + "num_input_tokens_seen": 9553032, + "step": 35630 + }, + { + "epoch": 9.34076015727392, + "grad_norm": 1.5683414936065674, + "learning_rate": 6.595353449146108e-07, + "loss": 0.2465, + "num_input_tokens_seen": 9554520, + "step": 35635 + }, + { + "epoch": 9.342070773263433, + "grad_norm": 1.6483954191207886, + "learning_rate": 6.569281147288414e-07, + "loss": 0.2601, + "num_input_tokens_seen": 9556424, + "step": 35640 + }, + { + "epoch": 9.343381389252949, + "grad_norm": 1.6148968935012817, + "learning_rate": 6.54325979505957e-07, + "loss": 0.1902, + "num_input_tokens_seen": 9558392, + "step": 35645 + }, + { + "epoch": 9.344692005242464, + "grad_norm": 5.987281799316406, + "learning_rate": 6.517289397905746e-07, + "loss": 0.1527, + "num_input_tokens_seen": 9559704, + "step": 35650 + }, + { + "epoch": 9.346002621231978, + "grad_norm": 2.068922519683838, + "learning_rate": 6.491369961262566e-07, + "loss": 0.2386, + "num_input_tokens_seen": 9561288, + "step": 35655 + }, + { + "epoch": 9.347313237221494, + "grad_norm": 3.7858211994171143, + "learning_rate": 6.465501490554914e-07, + "loss": 0.241, + "num_input_tokens_seen": 9562440, + "step": 35660 + }, + { + "epoch": 9.34862385321101, + "grad_norm": 2.0529332160949707, + "learning_rate": 6.439683991197043e-07, + "loss": 0.1107, + "num_input_tokens_seen": 9563880, + "step": 35665 + }, + { + "epoch": 9.349934469200525, + "grad_norm": 1.9774634838104248, + "learning_rate": 6.413917468592434e-07, + "loss": 0.1941, + "num_input_tokens_seen": 9565032, + "step": 35670 + }, + { + "epoch": 9.351245085190039, + "grad_norm": 1.941948652267456, + "learning_rate": 6.388201928134025e-07, + "loss": 0.0785, + "num_input_tokens_seen": 9566008, + "step": 35675 + }, + { + "epoch": 9.352555701179554, + "grad_norm": 1.7495166063308716, + "learning_rate": 6.362537375204009e-07, + "loss": 0.2356, + "num_input_tokens_seen": 9567336, + "step": 35680 + }, + { + "epoch": 9.35386631716907, + "grad_norm": 2.889598846435547, + "learning_rate": 6.336923815173951e-07, + "loss": 0.1099, + "num_input_tokens_seen": 9568312, + "step": 35685 + }, + { + "epoch": 9.355176933158585, + "grad_norm": 4.123291015625, + "learning_rate": 6.31136125340473e-07, + "loss": 0.1895, + "num_input_tokens_seen": 9569816, + "step": 35690 + }, + { + "epoch": 9.356487549148099, + "grad_norm": 3.806121826171875, + "learning_rate": 6.285849695246537e-07, + "loss": 0.1158, + "num_input_tokens_seen": 9571224, + "step": 35695 + }, + { + "epoch": 9.357798165137615, + "grad_norm": 0.4222785234451294, + "learning_rate": 6.260389146038881e-07, + "loss": 0.1519, + "num_input_tokens_seen": 9572360, + "step": 35700 + }, + { + "epoch": 9.35910878112713, + "grad_norm": 0.3814763128757477, + "learning_rate": 6.234979611110637e-07, + "loss": 0.1364, + "num_input_tokens_seen": 9573656, + "step": 35705 + }, + { + "epoch": 9.360419397116646, + "grad_norm": 2.4839863777160645, + "learning_rate": 6.209621095779993e-07, + "loss": 0.0783, + "num_input_tokens_seen": 9574776, + "step": 35710 + }, + { + "epoch": 9.36173001310616, + "grad_norm": 2.7195677757263184, + "learning_rate": 6.184313605354375e-07, + "loss": 0.1446, + "num_input_tokens_seen": 9576248, + "step": 35715 + }, + { + "epoch": 9.363040629095675, + "grad_norm": 2.062800407409668, + "learning_rate": 6.159057145130653e-07, + "loss": 0.1809, + "num_input_tokens_seen": 9577432, + "step": 35720 + }, + { + "epoch": 9.36435124508519, + "grad_norm": 2.449760913848877, + "learning_rate": 6.133851720394962e-07, + "loss": 0.0355, + "num_input_tokens_seen": 9578552, + "step": 35725 + }, + { + "epoch": 9.365661861074706, + "grad_norm": 1.793300986289978, + "learning_rate": 6.10869733642272e-07, + "loss": 0.2607, + "num_input_tokens_seen": 9580024, + "step": 35730 + }, + { + "epoch": 9.36697247706422, + "grad_norm": 0.7396408915519714, + "learning_rate": 6.083593998478715e-07, + "loss": 0.122, + "num_input_tokens_seen": 9581496, + "step": 35735 + }, + { + "epoch": 9.368283093053735, + "grad_norm": 4.000816345214844, + "learning_rate": 6.058541711817023e-07, + "loss": 0.1386, + "num_input_tokens_seen": 9582440, + "step": 35740 + }, + { + "epoch": 9.36959370904325, + "grad_norm": 1.3157672882080078, + "learning_rate": 6.033540481681088e-07, + "loss": 0.0965, + "num_input_tokens_seen": 9583896, + "step": 35745 + }, + { + "epoch": 9.370904325032765, + "grad_norm": 0.7970879077911377, + "learning_rate": 6.008590313303586e-07, + "loss": 0.1616, + "num_input_tokens_seen": 9585304, + "step": 35750 + }, + { + "epoch": 9.37221494102228, + "grad_norm": 2.22411847114563, + "learning_rate": 5.983691211906506e-07, + "loss": 0.1953, + "num_input_tokens_seen": 9586488, + "step": 35755 + }, + { + "epoch": 9.373525557011796, + "grad_norm": 3.5218076705932617, + "learning_rate": 5.958843182701263e-07, + "loss": 0.1861, + "num_input_tokens_seen": 9587880, + "step": 35760 + }, + { + "epoch": 9.374836173001311, + "grad_norm": 2.9560019969940186, + "learning_rate": 5.934046230888418e-07, + "loss": 0.1416, + "num_input_tokens_seen": 9588920, + "step": 35765 + }, + { + "epoch": 9.376146788990825, + "grad_norm": 4.6007232666015625, + "learning_rate": 5.909300361658015e-07, + "loss": 0.1225, + "num_input_tokens_seen": 9589976, + "step": 35770 + }, + { + "epoch": 9.37745740498034, + "grad_norm": 1.1398537158966064, + "learning_rate": 5.8846055801893e-07, + "loss": 0.1095, + "num_input_tokens_seen": 9591704, + "step": 35775 + }, + { + "epoch": 9.378768020969856, + "grad_norm": 3.2338831424713135, + "learning_rate": 5.859961891650861e-07, + "loss": 0.1983, + "num_input_tokens_seen": 9593000, + "step": 35780 + }, + { + "epoch": 9.380078636959372, + "grad_norm": 3.602674722671509, + "learning_rate": 5.835369301200516e-07, + "loss": 0.1489, + "num_input_tokens_seen": 9594440, + "step": 35785 + }, + { + "epoch": 9.381389252948885, + "grad_norm": 2.033186435699463, + "learning_rate": 5.810827813985509e-07, + "loss": 0.0991, + "num_input_tokens_seen": 9595672, + "step": 35790 + }, + { + "epoch": 9.382699868938401, + "grad_norm": 2.022305488586426, + "learning_rate": 5.786337435142314e-07, + "loss": 0.143, + "num_input_tokens_seen": 9597112, + "step": 35795 + }, + { + "epoch": 9.384010484927916, + "grad_norm": 0.00461959931999445, + "learning_rate": 5.761898169796721e-07, + "loss": 0.0599, + "num_input_tokens_seen": 9598456, + "step": 35800 + }, + { + "epoch": 9.385321100917432, + "grad_norm": 2.8802127838134766, + "learning_rate": 5.73751002306383e-07, + "loss": 0.2225, + "num_input_tokens_seen": 9599608, + "step": 35805 + }, + { + "epoch": 9.386631716906946, + "grad_norm": 2.102634906768799, + "learning_rate": 5.713173000048033e-07, + "loss": 0.1287, + "num_input_tokens_seen": 9601000, + "step": 35810 + }, + { + "epoch": 9.387942332896461, + "grad_norm": 1.2370554208755493, + "learning_rate": 5.688887105843032e-07, + "loss": 0.1385, + "num_input_tokens_seen": 9602232, + "step": 35815 + }, + { + "epoch": 9.389252948885977, + "grad_norm": 2.718101978302002, + "learning_rate": 5.664652345531846e-07, + "loss": 0.1199, + "num_input_tokens_seen": 9603400, + "step": 35820 + }, + { + "epoch": 9.390563564875492, + "grad_norm": 1.506805419921875, + "learning_rate": 5.640468724186721e-07, + "loss": 0.1621, + "num_input_tokens_seen": 9604648, + "step": 35825 + }, + { + "epoch": 9.391874180865006, + "grad_norm": 2.304368019104004, + "learning_rate": 5.616336246869275e-07, + "loss": 0.0855, + "num_input_tokens_seen": 9606008, + "step": 35830 + }, + { + "epoch": 9.393184796854522, + "grad_norm": 2.4238157272338867, + "learning_rate": 5.592254918630385e-07, + "loss": 0.2174, + "num_input_tokens_seen": 9607304, + "step": 35835 + }, + { + "epoch": 9.394495412844037, + "grad_norm": 5.4075188636779785, + "learning_rate": 5.568224744510242e-07, + "loss": 0.175, + "num_input_tokens_seen": 9608360, + "step": 35840 + }, + { + "epoch": 9.395806028833551, + "grad_norm": 1.2060359716415405, + "learning_rate": 5.544245729538322e-07, + "loss": 0.1677, + "num_input_tokens_seen": 9609720, + "step": 35845 + }, + { + "epoch": 9.397116644823067, + "grad_norm": 3.7570977210998535, + "learning_rate": 5.520317878733389e-07, + "loss": 0.1027, + "num_input_tokens_seen": 9610536, + "step": 35850 + }, + { + "epoch": 9.398427260812582, + "grad_norm": 1.8829288482666016, + "learning_rate": 5.496441197103463e-07, + "loss": 0.1582, + "num_input_tokens_seen": 9611832, + "step": 35855 + }, + { + "epoch": 9.399737876802098, + "grad_norm": 1.404221534729004, + "learning_rate": 5.472615689645965e-07, + "loss": 0.1514, + "num_input_tokens_seen": 9613416, + "step": 35860 + }, + { + "epoch": 9.401048492791611, + "grad_norm": 1.5454695224761963, + "learning_rate": 5.44884136134749e-07, + "loss": 0.1241, + "num_input_tokens_seen": 9614296, + "step": 35865 + }, + { + "epoch": 9.402359108781127, + "grad_norm": 1.7811987400054932, + "learning_rate": 5.425118217183945e-07, + "loss": 0.0316, + "num_input_tokens_seen": 9615176, + "step": 35870 + }, + { + "epoch": 9.403669724770642, + "grad_norm": 2.181853771209717, + "learning_rate": 5.40144626212058e-07, + "loss": 0.1291, + "num_input_tokens_seen": 9616472, + "step": 35875 + }, + { + "epoch": 9.404980340760158, + "grad_norm": 1.7681087255477905, + "learning_rate": 5.377825501111849e-07, + "loss": 0.1182, + "num_input_tokens_seen": 9617784, + "step": 35880 + }, + { + "epoch": 9.406290956749672, + "grad_norm": 1.4248745441436768, + "learning_rate": 5.354255939101576e-07, + "loss": 0.1363, + "num_input_tokens_seen": 9619272, + "step": 35885 + }, + { + "epoch": 9.407601572739187, + "grad_norm": 2.577345609664917, + "learning_rate": 5.330737581022843e-07, + "loss": 0.1398, + "num_input_tokens_seen": 9620472, + "step": 35890 + }, + { + "epoch": 9.408912188728703, + "grad_norm": 0.3393584191799164, + "learning_rate": 5.307270431797962e-07, + "loss": 0.1369, + "num_input_tokens_seen": 9622040, + "step": 35895 + }, + { + "epoch": 9.410222804718218, + "grad_norm": 2.1934001445770264, + "learning_rate": 5.283854496338586e-07, + "loss": 0.2029, + "num_input_tokens_seen": 9623304, + "step": 35900 + }, + { + "epoch": 9.411533420707732, + "grad_norm": 1.5504260063171387, + "learning_rate": 5.26048977954563e-07, + "loss": 0.103, + "num_input_tokens_seen": 9624936, + "step": 35905 + }, + { + "epoch": 9.412844036697248, + "grad_norm": 1.805187702178955, + "learning_rate": 5.237176286309292e-07, + "loss": 0.1052, + "num_input_tokens_seen": 9626952, + "step": 35910 + }, + { + "epoch": 9.414154652686763, + "grad_norm": 2.551349639892578, + "learning_rate": 5.213914021509031e-07, + "loss": 0.2003, + "num_input_tokens_seen": 9628200, + "step": 35915 + }, + { + "epoch": 9.415465268676279, + "grad_norm": 0.788345456123352, + "learning_rate": 5.190702990013618e-07, + "loss": 0.1197, + "num_input_tokens_seen": 9629432, + "step": 35920 + }, + { + "epoch": 9.416775884665793, + "grad_norm": 2.336848020553589, + "learning_rate": 5.167543196681058e-07, + "loss": 0.0686, + "num_input_tokens_seen": 9630824, + "step": 35925 + }, + { + "epoch": 9.418086500655308, + "grad_norm": 1.8957160711288452, + "learning_rate": 5.144434646358665e-07, + "loss": 0.1751, + "num_input_tokens_seen": 9632104, + "step": 35930 + }, + { + "epoch": 9.419397116644824, + "grad_norm": 1.590137004852295, + "learning_rate": 5.121377343883016e-07, + "loss": 0.1148, + "num_input_tokens_seen": 9633704, + "step": 35935 + }, + { + "epoch": 9.420707732634337, + "grad_norm": 1.107006311416626, + "learning_rate": 5.098371294080001e-07, + "loss": 0.0739, + "num_input_tokens_seen": 9634968, + "step": 35940 + }, + { + "epoch": 9.422018348623853, + "grad_norm": 2.4013471603393555, + "learning_rate": 5.075416501764685e-07, + "loss": 0.3134, + "num_input_tokens_seen": 9636392, + "step": 35945 + }, + { + "epoch": 9.423328964613368, + "grad_norm": 1.9595547914505005, + "learning_rate": 5.052512971741529e-07, + "loss": 0.1515, + "num_input_tokens_seen": 9637400, + "step": 35950 + }, + { + "epoch": 9.424639580602884, + "grad_norm": 1.8731220960617065, + "learning_rate": 5.029660708804146e-07, + "loss": 0.187, + "num_input_tokens_seen": 9638952, + "step": 35955 + }, + { + "epoch": 9.425950196592398, + "grad_norm": 3.0215659141540527, + "learning_rate": 5.006859717735512e-07, + "loss": 0.1651, + "num_input_tokens_seen": 9640232, + "step": 35960 + }, + { + "epoch": 9.427260812581913, + "grad_norm": 1.2237865924835205, + "learning_rate": 4.984110003307812e-07, + "loss": 0.1407, + "num_input_tokens_seen": 9641384, + "step": 35965 + }, + { + "epoch": 9.428571428571429, + "grad_norm": 1.471003532409668, + "learning_rate": 4.961411570282543e-07, + "loss": 0.0737, + "num_input_tokens_seen": 9643128, + "step": 35970 + }, + { + "epoch": 9.429882044560944, + "grad_norm": 1.6221787929534912, + "learning_rate": 4.938764423410458e-07, + "loss": 0.0803, + "num_input_tokens_seen": 9644312, + "step": 35975 + }, + { + "epoch": 9.431192660550458, + "grad_norm": 3.218275308609009, + "learning_rate": 4.916168567431544e-07, + "loss": 0.1394, + "num_input_tokens_seen": 9645544, + "step": 35980 + }, + { + "epoch": 9.432503276539974, + "grad_norm": 2.4232096672058105, + "learning_rate": 4.893624007075048e-07, + "loss": 0.1623, + "num_input_tokens_seen": 9646888, + "step": 35985 + }, + { + "epoch": 9.43381389252949, + "grad_norm": 2.34439754486084, + "learning_rate": 4.871130747059554e-07, + "loss": 0.1533, + "num_input_tokens_seen": 9648360, + "step": 35990 + }, + { + "epoch": 9.435124508519005, + "grad_norm": 2.3302226066589355, + "learning_rate": 4.848688792092798e-07, + "loss": 0.18, + "num_input_tokens_seen": 9649608, + "step": 35995 + }, + { + "epoch": 9.436435124508519, + "grad_norm": 1.1035139560699463, + "learning_rate": 4.826298146871938e-07, + "loss": 0.2426, + "num_input_tokens_seen": 9650840, + "step": 36000 + }, + { + "epoch": 9.437745740498034, + "grad_norm": 2.077476739883423, + "learning_rate": 4.803958816083254e-07, + "loss": 0.1293, + "num_input_tokens_seen": 9651928, + "step": 36005 + }, + { + "epoch": 9.43905635648755, + "grad_norm": 3.8220818042755127, + "learning_rate": 4.781670804402283e-07, + "loss": 0.185, + "num_input_tokens_seen": 9653192, + "step": 36010 + }, + { + "epoch": 9.440366972477065, + "grad_norm": 2.617293357849121, + "learning_rate": 4.75943411649396e-07, + "loss": 0.1001, + "num_input_tokens_seen": 9654152, + "step": 36015 + }, + { + "epoch": 9.441677588466579, + "grad_norm": 1.4862526655197144, + "learning_rate": 4.7372487570123135e-07, + "loss": 0.1797, + "num_input_tokens_seen": 9655656, + "step": 36020 + }, + { + "epoch": 9.442988204456094, + "grad_norm": 1.440673828125, + "learning_rate": 4.7151147306007117e-07, + "loss": 0.1283, + "num_input_tokens_seen": 9656792, + "step": 36025 + }, + { + "epoch": 9.44429882044561, + "grad_norm": 2.9325735569000244, + "learning_rate": 4.6930320418917817e-07, + "loss": 0.1466, + "num_input_tokens_seen": 9657928, + "step": 36030 + }, + { + "epoch": 9.445609436435124, + "grad_norm": 2.3451809883117676, + "learning_rate": 4.6710006955073826e-07, + "loss": 0.1225, + "num_input_tokens_seen": 9659032, + "step": 36035 + }, + { + "epoch": 9.44692005242464, + "grad_norm": 2.779128313064575, + "learning_rate": 4.649020696058659e-07, + "loss": 0.1422, + "num_input_tokens_seen": 9660200, + "step": 36040 + }, + { + "epoch": 9.448230668414155, + "grad_norm": 2.374788761138916, + "learning_rate": 4.627092048145959e-07, + "loss": 0.0701, + "num_input_tokens_seen": 9661432, + "step": 36045 + }, + { + "epoch": 9.44954128440367, + "grad_norm": 2.3943912982940674, + "learning_rate": 4.6052147563589167e-07, + "loss": 0.2109, + "num_input_tokens_seen": 9662776, + "step": 36050 + }, + { + "epoch": 9.450851900393184, + "grad_norm": 0.5963407754898071, + "learning_rate": 4.583388825276397e-07, + "loss": 0.1157, + "num_input_tokens_seen": 9663896, + "step": 36055 + }, + { + "epoch": 9.4521625163827, + "grad_norm": 1.152753472328186, + "learning_rate": 4.561614259466579e-07, + "loss": 0.0654, + "num_input_tokens_seen": 9665144, + "step": 36060 + }, + { + "epoch": 9.453473132372215, + "grad_norm": 3.8918044567108154, + "learning_rate": 4.53989106348679e-07, + "loss": 0.2837, + "num_input_tokens_seen": 9666616, + "step": 36065 + }, + { + "epoch": 9.45478374836173, + "grad_norm": 1.6851632595062256, + "learning_rate": 4.518219241883698e-07, + "loss": 0.0818, + "num_input_tokens_seen": 9667768, + "step": 36070 + }, + { + "epoch": 9.456094364351245, + "grad_norm": 1.1151069402694702, + "learning_rate": 4.4965987991931466e-07, + "loss": 0.0897, + "num_input_tokens_seen": 9669192, + "step": 36075 + }, + { + "epoch": 9.45740498034076, + "grad_norm": 0.7285088300704956, + "learning_rate": 4.475029739940295e-07, + "loss": 0.1121, + "num_input_tokens_seen": 9670392, + "step": 36080 + }, + { + "epoch": 9.458715596330276, + "grad_norm": 1.9380433559417725, + "learning_rate": 4.4535120686394484e-07, + "loss": 0.15, + "num_input_tokens_seen": 9671656, + "step": 36085 + }, + { + "epoch": 9.460026212319791, + "grad_norm": 0.7762132883071899, + "learning_rate": 4.4320457897942815e-07, + "loss": 0.0589, + "num_input_tokens_seen": 9672696, + "step": 36090 + }, + { + "epoch": 9.461336828309305, + "grad_norm": 0.8198335766792297, + "learning_rate": 4.410630907897617e-07, + "loss": 0.1427, + "num_input_tokens_seen": 9674152, + "step": 36095 + }, + { + "epoch": 9.46264744429882, + "grad_norm": 1.615822434425354, + "learning_rate": 4.389267427431565e-07, + "loss": 0.0785, + "num_input_tokens_seen": 9675272, + "step": 36100 + }, + { + "epoch": 9.463958060288336, + "grad_norm": 1.6421406269073486, + "learning_rate": 4.367955352867437e-07, + "loss": 0.1514, + "num_input_tokens_seen": 9676600, + "step": 36105 + }, + { + "epoch": 9.46526867627785, + "grad_norm": 3.573791742324829, + "learning_rate": 4.346694688665831e-07, + "loss": 0.1248, + "num_input_tokens_seen": 9677656, + "step": 36110 + }, + { + "epoch": 9.466579292267365, + "grad_norm": 2.8221449851989746, + "learning_rate": 4.325485439276578e-07, + "loss": 0.0945, + "num_input_tokens_seen": 9678696, + "step": 36115 + }, + { + "epoch": 9.46788990825688, + "grad_norm": 1.4726066589355469, + "learning_rate": 4.304327609138736e-07, + "loss": 0.1146, + "num_input_tokens_seen": 9680280, + "step": 36120 + }, + { + "epoch": 9.469200524246396, + "grad_norm": 1.6395236253738403, + "learning_rate": 4.2832212026805986e-07, + "loss": 0.1548, + "num_input_tokens_seen": 9681432, + "step": 36125 + }, + { + "epoch": 9.47051114023591, + "grad_norm": 4.241192817687988, + "learning_rate": 4.262166224319686e-07, + "loss": 0.1637, + "num_input_tokens_seen": 9682744, + "step": 36130 + }, + { + "epoch": 9.471821756225426, + "grad_norm": 1.1178749799728394, + "learning_rate": 4.2411626784628065e-07, + "loss": 0.2583, + "num_input_tokens_seen": 9686088, + "step": 36135 + }, + { + "epoch": 9.473132372214941, + "grad_norm": 2.196347951889038, + "learning_rate": 4.220210569505917e-07, + "loss": 0.1337, + "num_input_tokens_seen": 9687240, + "step": 36140 + }, + { + "epoch": 9.474442988204457, + "grad_norm": 0.5358837246894836, + "learning_rate": 4.199309901834286e-07, + "loss": 0.1393, + "num_input_tokens_seen": 9688440, + "step": 36145 + }, + { + "epoch": 9.47575360419397, + "grad_norm": 1.4334355592727661, + "learning_rate": 4.178460679822388e-07, + "loss": 0.209, + "num_input_tokens_seen": 9690136, + "step": 36150 + }, + { + "epoch": 9.477064220183486, + "grad_norm": 2.440445899963379, + "learning_rate": 4.157662907833898e-07, + "loss": 0.1014, + "num_input_tokens_seen": 9691352, + "step": 36155 + }, + { + "epoch": 9.478374836173002, + "grad_norm": 1.3952101469039917, + "learning_rate": 4.1369165902218065e-07, + "loss": 0.216, + "num_input_tokens_seen": 9692760, + "step": 36160 + }, + { + "epoch": 9.479685452162517, + "grad_norm": 1.3605310916900635, + "learning_rate": 4.1162217313282236e-07, + "loss": 0.1958, + "num_input_tokens_seen": 9694024, + "step": 36165 + }, + { + "epoch": 9.48099606815203, + "grad_norm": 1.2046397924423218, + "learning_rate": 4.0955783354846013e-07, + "loss": 0.1257, + "num_input_tokens_seen": 9695192, + "step": 36170 + }, + { + "epoch": 9.482306684141546, + "grad_norm": 1.4147698879241943, + "learning_rate": 4.0749864070115387e-07, + "loss": 0.1843, + "num_input_tokens_seen": 9696536, + "step": 36175 + }, + { + "epoch": 9.483617300131062, + "grad_norm": 3.7870187759399414, + "learning_rate": 4.0544459502188934e-07, + "loss": 0.0827, + "num_input_tokens_seen": 9697928, + "step": 36180 + }, + { + "epoch": 9.484927916120578, + "grad_norm": 2.0227482318878174, + "learning_rate": 4.0339569694057554e-07, + "loss": 0.1736, + "num_input_tokens_seen": 9699272, + "step": 36185 + }, + { + "epoch": 9.486238532110091, + "grad_norm": 2.318895101547241, + "learning_rate": 4.013519468860416e-07, + "loss": 0.1764, + "num_input_tokens_seen": 9700968, + "step": 36190 + }, + { + "epoch": 9.487549148099607, + "grad_norm": 1.739776372909546, + "learning_rate": 3.993133452860426e-07, + "loss": 0.1999, + "num_input_tokens_seen": 9702344, + "step": 36195 + }, + { + "epoch": 9.488859764089122, + "grad_norm": 2.97204852104187, + "learning_rate": 3.9727989256725384e-07, + "loss": 0.1226, + "num_input_tokens_seen": 9703352, + "step": 36200 + }, + { + "epoch": 9.490170380078636, + "grad_norm": 0.6563434600830078, + "learning_rate": 3.9525158915527383e-07, + "loss": 0.1424, + "num_input_tokens_seen": 9704840, + "step": 36205 + }, + { + "epoch": 9.491480996068152, + "grad_norm": 2.759523391723633, + "learning_rate": 3.932284354746213e-07, + "loss": 0.1238, + "num_input_tokens_seen": 9706248, + "step": 36210 + }, + { + "epoch": 9.492791612057667, + "grad_norm": 4.5310211181640625, + "learning_rate": 3.9121043194874365e-07, + "loss": 0.113, + "num_input_tokens_seen": 9707384, + "step": 36215 + }, + { + "epoch": 9.494102228047183, + "grad_norm": 0.3575455844402313, + "learning_rate": 3.8919757899999743e-07, + "loss": 0.0964, + "num_input_tokens_seen": 9709208, + "step": 36220 + }, + { + "epoch": 9.495412844036696, + "grad_norm": 1.8993569612503052, + "learning_rate": 3.8718987704967345e-07, + "loss": 0.1366, + "num_input_tokens_seen": 9710392, + "step": 36225 + }, + { + "epoch": 9.496723460026212, + "grad_norm": 6.487475872039795, + "learning_rate": 3.8518732651798284e-07, + "loss": 0.198, + "num_input_tokens_seen": 9711624, + "step": 36230 + }, + { + "epoch": 9.498034076015728, + "grad_norm": 0.943873941898346, + "learning_rate": 3.8318992782405414e-07, + "loss": 0.1174, + "num_input_tokens_seen": 9712680, + "step": 36235 + }, + { + "epoch": 9.499344692005243, + "grad_norm": 1.392050862312317, + "learning_rate": 3.811976813859364e-07, + "loss": 0.0921, + "num_input_tokens_seen": 9714008, + "step": 36240 + }, + { + "epoch": 9.500655307994757, + "grad_norm": 1.0969688892364502, + "learning_rate": 3.792105876206098e-07, + "loss": 0.1541, + "num_input_tokens_seen": 9715176, + "step": 36245 + }, + { + "epoch": 9.501965923984272, + "grad_norm": 3.7279560565948486, + "learning_rate": 3.7722864694396397e-07, + "loss": 0.1177, + "num_input_tokens_seen": 9716296, + "step": 36250 + }, + { + "epoch": 9.502490170380078, + "eval_loss": 0.8772965669631958, + "eval_runtime": 16.8533, + "eval_samples_per_second": 50.317, + "eval_steps_per_second": 25.158, + "num_input_tokens_seen": 9716920, + "step": 36252 + }, + { + "epoch": 9.503276539973788, + "grad_norm": 0.7206158638000488, + "learning_rate": 3.7525185977081713e-07, + "loss": 0.1199, + "num_input_tokens_seen": 9719064, + "step": 36255 + }, + { + "epoch": 9.504587155963304, + "grad_norm": 1.4070791006088257, + "learning_rate": 3.7328022651490767e-07, + "loss": 0.1425, + "num_input_tokens_seen": 9720472, + "step": 36260 + }, + { + "epoch": 9.505897771952817, + "grad_norm": 3.016559600830078, + "learning_rate": 3.7131374758889446e-07, + "loss": 0.1215, + "num_input_tokens_seen": 9721896, + "step": 36265 + }, + { + "epoch": 9.507208387942333, + "grad_norm": 0.7828986644744873, + "learning_rate": 3.693524234043594e-07, + "loss": 0.1738, + "num_input_tokens_seen": 9723480, + "step": 36270 + }, + { + "epoch": 9.508519003931848, + "grad_norm": 0.8672478199005127, + "learning_rate": 3.6739625437180457e-07, + "loss": 0.1459, + "num_input_tokens_seen": 9725224, + "step": 36275 + }, + { + "epoch": 9.509829619921362, + "grad_norm": 4.626498222351074, + "learning_rate": 3.654452409006498e-07, + "loss": 0.3996, + "num_input_tokens_seen": 9726520, + "step": 36280 + }, + { + "epoch": 9.511140235910878, + "grad_norm": 2.166022539138794, + "learning_rate": 3.634993833992434e-07, + "loss": 0.0894, + "num_input_tokens_seen": 9727352, + "step": 36285 + }, + { + "epoch": 9.512450851900393, + "grad_norm": 3.1443488597869873, + "learning_rate": 3.615586822748457e-07, + "loss": 0.0975, + "num_input_tokens_seen": 9728760, + "step": 36290 + }, + { + "epoch": 9.513761467889909, + "grad_norm": 0.8902989625930786, + "learning_rate": 3.596231379336429e-07, + "loss": 0.0946, + "num_input_tokens_seen": 9730360, + "step": 36295 + }, + { + "epoch": 9.515072083879424, + "grad_norm": 3.736685276031494, + "learning_rate": 3.5769275078074425e-07, + "loss": 0.1334, + "num_input_tokens_seen": 9731688, + "step": 36300 + }, + { + "epoch": 9.516382699868938, + "grad_norm": 1.3244754076004028, + "learning_rate": 3.5576752122017385e-07, + "loss": 0.2273, + "num_input_tokens_seen": 9733160, + "step": 36305 + }, + { + "epoch": 9.517693315858454, + "grad_norm": 0.18856993317604065, + "learning_rate": 3.5384744965487883e-07, + "loss": 0.0992, + "num_input_tokens_seen": 9734760, + "step": 36310 + }, + { + "epoch": 9.51900393184797, + "grad_norm": 1.7087867259979248, + "learning_rate": 3.519325364867265e-07, + "loss": 0.0892, + "num_input_tokens_seen": 9735960, + "step": 36315 + }, + { + "epoch": 9.520314547837483, + "grad_norm": 0.5614672303199768, + "learning_rate": 3.5002278211650476e-07, + "loss": 0.1743, + "num_input_tokens_seen": 9737416, + "step": 36320 + }, + { + "epoch": 9.521625163826998, + "grad_norm": 1.2520580291748047, + "learning_rate": 3.4811818694392706e-07, + "loss": 0.1689, + "num_input_tokens_seen": 9738888, + "step": 36325 + }, + { + "epoch": 9.522935779816514, + "grad_norm": 6.708319187164307, + "learning_rate": 3.462187513676135e-07, + "loss": 0.0643, + "num_input_tokens_seen": 9739880, + "step": 36330 + }, + { + "epoch": 9.52424639580603, + "grad_norm": 0.984679639339447, + "learning_rate": 3.4432447578511826e-07, + "loss": 0.0937, + "num_input_tokens_seen": 9741336, + "step": 36335 + }, + { + "epoch": 9.525557011795543, + "grad_norm": 1.6490983963012695, + "learning_rate": 3.4243536059291035e-07, + "loss": 0.1941, + "num_input_tokens_seen": 9742776, + "step": 36340 + }, + { + "epoch": 9.526867627785059, + "grad_norm": 1.8571767807006836, + "learning_rate": 3.405514061863735e-07, + "loss": 0.0908, + "num_input_tokens_seen": 9744072, + "step": 36345 + }, + { + "epoch": 9.528178243774574, + "grad_norm": 2.9942877292633057, + "learning_rate": 3.3867261295982e-07, + "loss": 0.2079, + "num_input_tokens_seen": 9745384, + "step": 36350 + }, + { + "epoch": 9.52948885976409, + "grad_norm": 1.7650541067123413, + "learning_rate": 3.36798981306477e-07, + "loss": 0.2697, + "num_input_tokens_seen": 9747400, + "step": 36355 + }, + { + "epoch": 9.530799475753604, + "grad_norm": 1.9080805778503418, + "learning_rate": 3.3493051161849477e-07, + "loss": 0.1744, + "num_input_tokens_seen": 9749032, + "step": 36360 + }, + { + "epoch": 9.53211009174312, + "grad_norm": 1.792994499206543, + "learning_rate": 3.330672042869354e-07, + "loss": 0.195, + "num_input_tokens_seen": 9750648, + "step": 36365 + }, + { + "epoch": 9.533420707732635, + "grad_norm": 3.1732802391052246, + "learning_rate": 3.312090597017925e-07, + "loss": 0.1156, + "num_input_tokens_seen": 9752392, + "step": 36370 + }, + { + "epoch": 9.534731323722148, + "grad_norm": 1.6907494068145752, + "learning_rate": 3.2935607825196614e-07, + "loss": 0.1068, + "num_input_tokens_seen": 9754152, + "step": 36375 + }, + { + "epoch": 9.536041939711664, + "grad_norm": 1.9418044090270996, + "learning_rate": 3.275082603252877e-07, + "loss": 0.0785, + "num_input_tokens_seen": 9755128, + "step": 36380 + }, + { + "epoch": 9.53735255570118, + "grad_norm": 1.5768301486968994, + "learning_rate": 3.2566560630849784e-07, + "loss": 0.0825, + "num_input_tokens_seen": 9756408, + "step": 36385 + }, + { + "epoch": 9.538663171690695, + "grad_norm": 4.478271961212158, + "learning_rate": 3.2382811658726033e-07, + "loss": 0.1829, + "num_input_tokens_seen": 9757432, + "step": 36390 + }, + { + "epoch": 9.539973787680209, + "grad_norm": 1.1403067111968994, + "learning_rate": 3.219957915461619e-07, + "loss": 0.0727, + "num_input_tokens_seen": 9758856, + "step": 36395 + }, + { + "epoch": 9.541284403669724, + "grad_norm": 0.7559523582458496, + "learning_rate": 3.201686315687041e-07, + "loss": 0.1351, + "num_input_tokens_seen": 9761496, + "step": 36400 + }, + { + "epoch": 9.54259501965924, + "grad_norm": 1.6590486764907837, + "learning_rate": 3.183466370373117e-07, + "loss": 0.145, + "num_input_tokens_seen": 9762616, + "step": 36405 + }, + { + "epoch": 9.543905635648755, + "grad_norm": 3.6923346519470215, + "learning_rate": 3.1652980833331846e-07, + "loss": 0.1571, + "num_input_tokens_seen": 9763864, + "step": 36410 + }, + { + "epoch": 9.54521625163827, + "grad_norm": 2.308030128479004, + "learning_rate": 3.147181458369897e-07, + "loss": 0.1474, + "num_input_tokens_seen": 9764968, + "step": 36415 + }, + { + "epoch": 9.546526867627785, + "grad_norm": 4.121222972869873, + "learning_rate": 3.12911649927497e-07, + "loss": 0.1954, + "num_input_tokens_seen": 9766424, + "step": 36420 + }, + { + "epoch": 9.5478374836173, + "grad_norm": 1.7996201515197754, + "learning_rate": 3.111103209829408e-07, + "loss": 0.0972, + "num_input_tokens_seen": 9767912, + "step": 36425 + }, + { + "epoch": 9.549148099606816, + "grad_norm": 5.776588439941406, + "learning_rate": 3.0931415938033336e-07, + "loss": 0.1082, + "num_input_tokens_seen": 9768904, + "step": 36430 + }, + { + "epoch": 9.55045871559633, + "grad_norm": 1.8195457458496094, + "learning_rate": 3.075231654956129e-07, + "loss": 0.1571, + "num_input_tokens_seen": 9770552, + "step": 36435 + }, + { + "epoch": 9.551769331585845, + "grad_norm": 2.896151542663574, + "learning_rate": 3.0573733970362674e-07, + "loss": 0.2277, + "num_input_tokens_seen": 9771704, + "step": 36440 + }, + { + "epoch": 9.55307994757536, + "grad_norm": 0.8981448411941528, + "learning_rate": 3.0395668237814813e-07, + "loss": 0.1817, + "num_input_tokens_seen": 9773320, + "step": 36445 + }, + { + "epoch": 9.554390563564876, + "grad_norm": 0.2967493236064911, + "learning_rate": 3.0218119389186503e-07, + "loss": 0.1473, + "num_input_tokens_seen": 9774424, + "step": 36450 + }, + { + "epoch": 9.55570117955439, + "grad_norm": 1.8911420106887817, + "learning_rate": 3.0041087461638304e-07, + "loss": 0.0602, + "num_input_tokens_seen": 9775336, + "step": 36455 + }, + { + "epoch": 9.557011795543906, + "grad_norm": 2.229337692260742, + "learning_rate": 2.986457249222252e-07, + "loss": 0.2043, + "num_input_tokens_seen": 9776680, + "step": 36460 + }, + { + "epoch": 9.558322411533421, + "grad_norm": 1.667380928993225, + "learning_rate": 2.968857451788404e-07, + "loss": 0.1639, + "num_input_tokens_seen": 9778088, + "step": 36465 + }, + { + "epoch": 9.559633027522935, + "grad_norm": 1.4248557090759277, + "learning_rate": 2.951309357545812e-07, + "loss": 0.1169, + "num_input_tokens_seen": 9779752, + "step": 36470 + }, + { + "epoch": 9.56094364351245, + "grad_norm": 1.2314543724060059, + "learning_rate": 2.9338129701673434e-07, + "loss": 0.1098, + "num_input_tokens_seen": 9781368, + "step": 36475 + }, + { + "epoch": 9.562254259501966, + "grad_norm": 0.9082302451133728, + "learning_rate": 2.916368293314903e-07, + "loss": 0.0593, + "num_input_tokens_seen": 9782600, + "step": 36480 + }, + { + "epoch": 9.563564875491481, + "grad_norm": 3.528637170791626, + "learning_rate": 2.898975330639653e-07, + "loss": 0.157, + "num_input_tokens_seen": 9784184, + "step": 36485 + }, + { + "epoch": 9.564875491480995, + "grad_norm": 1.9071677923202515, + "learning_rate": 2.8816340857819034e-07, + "loss": 0.1182, + "num_input_tokens_seen": 9785672, + "step": 36490 + }, + { + "epoch": 9.56618610747051, + "grad_norm": 1.6227132081985474, + "learning_rate": 2.8643445623711684e-07, + "loss": 0.1804, + "num_input_tokens_seen": 9786952, + "step": 36495 + }, + { + "epoch": 9.567496723460026, + "grad_norm": 1.2198474407196045, + "learning_rate": 2.847106764026081e-07, + "loss": 0.0912, + "num_input_tokens_seen": 9788040, + "step": 36500 + }, + { + "epoch": 9.568807339449542, + "grad_norm": 3.728006601333618, + "learning_rate": 2.8299206943545044e-07, + "loss": 0.1348, + "num_input_tokens_seen": 9789144, + "step": 36505 + }, + { + "epoch": 9.570117955439056, + "grad_norm": 2.8270649909973145, + "learning_rate": 2.812786356953451e-07, + "loss": 0.0902, + "num_input_tokens_seen": 9790168, + "step": 36510 + }, + { + "epoch": 9.571428571428571, + "grad_norm": 1.4357830286026, + "learning_rate": 2.795703755409107e-07, + "loss": 0.1586, + "num_input_tokens_seen": 9791672, + "step": 36515 + }, + { + "epoch": 9.572739187418087, + "grad_norm": 1.5419244766235352, + "learning_rate": 2.7786728932968354e-07, + "loss": 0.1582, + "num_input_tokens_seen": 9793048, + "step": 36520 + }, + { + "epoch": 9.574049803407602, + "grad_norm": 1.8591532707214355, + "learning_rate": 2.761693774181173e-07, + "loss": 0.1955, + "num_input_tokens_seen": 9794648, + "step": 36525 + }, + { + "epoch": 9.575360419397116, + "grad_norm": 0.7707093954086304, + "learning_rate": 2.7447664016157783e-07, + "loss": 0.0856, + "num_input_tokens_seen": 9795928, + "step": 36530 + }, + { + "epoch": 9.576671035386632, + "grad_norm": 3.162853956222534, + "learning_rate": 2.727890779143566e-07, + "loss": 0.1835, + "num_input_tokens_seen": 9797064, + "step": 36535 + }, + { + "epoch": 9.577981651376147, + "grad_norm": 1.1217732429504395, + "learning_rate": 2.7110669102965723e-07, + "loss": 0.1122, + "num_input_tokens_seen": 9798120, + "step": 36540 + }, + { + "epoch": 9.579292267365663, + "grad_norm": 1.7521015405654907, + "learning_rate": 2.694294798595981e-07, + "loss": 0.1762, + "num_input_tokens_seen": 9799336, + "step": 36545 + }, + { + "epoch": 9.580602883355176, + "grad_norm": 0.4132865369319916, + "learning_rate": 2.677574447552178e-07, + "loss": 0.1489, + "num_input_tokens_seen": 9800664, + "step": 36550 + }, + { + "epoch": 9.581913499344692, + "grad_norm": 3.4394328594207764, + "learning_rate": 2.660905860664697e-07, + "loss": 0.0675, + "num_input_tokens_seen": 9801656, + "step": 36555 + }, + { + "epoch": 9.583224115334207, + "grad_norm": 0.8786471486091614, + "learning_rate": 2.6442890414222487e-07, + "loss": 0.1212, + "num_input_tokens_seen": 9803416, + "step": 36560 + }, + { + "epoch": 9.584534731323721, + "grad_norm": 4.940483093261719, + "learning_rate": 2.6277239933027163e-07, + "loss": 0.1392, + "num_input_tokens_seen": 9804568, + "step": 36565 + }, + { + "epoch": 9.585845347313237, + "grad_norm": 2.4768431186676025, + "learning_rate": 2.611210719773133e-07, + "loss": 0.0681, + "num_input_tokens_seen": 9805528, + "step": 36570 + }, + { + "epoch": 9.587155963302752, + "grad_norm": 1.525214672088623, + "learning_rate": 2.5947492242896776e-07, + "loss": 0.0709, + "num_input_tokens_seen": 9806616, + "step": 36575 + }, + { + "epoch": 9.588466579292268, + "grad_norm": 2.9724133014678955, + "learning_rate": 2.5783395102977057e-07, + "loss": 0.178, + "num_input_tokens_seen": 9807944, + "step": 36580 + }, + { + "epoch": 9.589777195281782, + "grad_norm": 1.4400781393051147, + "learning_rate": 2.5619815812318037e-07, + "loss": 0.1877, + "num_input_tokens_seen": 9809272, + "step": 36585 + }, + { + "epoch": 9.591087811271297, + "grad_norm": 2.3434572219848633, + "learning_rate": 2.5456754405155934e-07, + "loss": 0.1827, + "num_input_tokens_seen": 9811080, + "step": 36590 + }, + { + "epoch": 9.592398427260813, + "grad_norm": 3.424558162689209, + "learning_rate": 2.5294210915619564e-07, + "loss": 0.1156, + "num_input_tokens_seen": 9811960, + "step": 36595 + }, + { + "epoch": 9.593709043250328, + "grad_norm": 1.1305887699127197, + "learning_rate": 2.5132185377728656e-07, + "loss": 0.0805, + "num_input_tokens_seen": 9813048, + "step": 36600 + }, + { + "epoch": 9.595019659239842, + "grad_norm": 2.6392006874084473, + "learning_rate": 2.497067782539553e-07, + "loss": 0.0964, + "num_input_tokens_seen": 9814168, + "step": 36605 + }, + { + "epoch": 9.596330275229358, + "grad_norm": 2.3668863773345947, + "learning_rate": 2.480968829242286e-07, + "loss": 0.0759, + "num_input_tokens_seen": 9815416, + "step": 36610 + }, + { + "epoch": 9.597640891218873, + "grad_norm": 1.8589054346084595, + "learning_rate": 2.4649216812505373e-07, + "loss": 0.129, + "num_input_tokens_seen": 9816632, + "step": 36615 + }, + { + "epoch": 9.598951507208389, + "grad_norm": 2.2370352745056152, + "learning_rate": 2.448926341923008e-07, + "loss": 0.0692, + "num_input_tokens_seen": 9817736, + "step": 36620 + }, + { + "epoch": 9.600262123197902, + "grad_norm": 1.4183613061904907, + "learning_rate": 2.4329828146074095e-07, + "loss": 0.1438, + "num_input_tokens_seen": 9819432, + "step": 36625 + }, + { + "epoch": 9.601572739187418, + "grad_norm": 4.590415000915527, + "learning_rate": 2.417091102640795e-07, + "loss": 0.1348, + "num_input_tokens_seen": 9821096, + "step": 36630 + }, + { + "epoch": 9.602883355176933, + "grad_norm": 2.2119667530059814, + "learning_rate": 2.401251209349198e-07, + "loss": 0.2069, + "num_input_tokens_seen": 9822584, + "step": 36635 + }, + { + "epoch": 9.604193971166449, + "grad_norm": 2.417778730392456, + "learning_rate": 2.385463138047911e-07, + "loss": 0.1228, + "num_input_tokens_seen": 9823720, + "step": 36640 + }, + { + "epoch": 9.605504587155963, + "grad_norm": 2.415398597717285, + "learning_rate": 2.3697268920413462e-07, + "loss": 0.1411, + "num_input_tokens_seen": 9824984, + "step": 36645 + }, + { + "epoch": 9.606815203145478, + "grad_norm": 1.9957059621810913, + "learning_rate": 2.3540424746230916e-07, + "loss": 0.3382, + "num_input_tokens_seen": 9826152, + "step": 36650 + }, + { + "epoch": 9.608125819134994, + "grad_norm": 3.3502302169799805, + "learning_rate": 2.3384098890758267e-07, + "loss": 0.0945, + "num_input_tokens_seen": 9827224, + "step": 36655 + }, + { + "epoch": 9.609436435124508, + "grad_norm": 0.2956158518791199, + "learning_rate": 2.322829138671434e-07, + "loss": 0.0874, + "num_input_tokens_seen": 9828472, + "step": 36660 + }, + { + "epoch": 9.610747051114023, + "grad_norm": 4.024384021759033, + "learning_rate": 2.3073002266709443e-07, + "loss": 0.1032, + "num_input_tokens_seen": 9829704, + "step": 36665 + }, + { + "epoch": 9.612057667103539, + "grad_norm": 2.17580509185791, + "learning_rate": 2.2918231563245629e-07, + "loss": 0.0851, + "num_input_tokens_seen": 9831224, + "step": 36670 + }, + { + "epoch": 9.613368283093054, + "grad_norm": 0.8769616484642029, + "learning_rate": 2.2763979308715877e-07, + "loss": 0.1299, + "num_input_tokens_seen": 9832744, + "step": 36675 + }, + { + "epoch": 9.614678899082568, + "grad_norm": 1.9720189571380615, + "learning_rate": 2.2610245535404638e-07, + "loss": 0.1377, + "num_input_tokens_seen": 9833960, + "step": 36680 + }, + { + "epoch": 9.615989515072084, + "grad_norm": 3.2778379917144775, + "learning_rate": 2.2457030275488676e-07, + "loss": 0.156, + "num_input_tokens_seen": 9835960, + "step": 36685 + }, + { + "epoch": 9.617300131061599, + "grad_norm": 3.2613186836242676, + "learning_rate": 2.2304333561035396e-07, + "loss": 0.1339, + "num_input_tokens_seen": 9836936, + "step": 36690 + }, + { + "epoch": 9.618610747051115, + "grad_norm": 4.349628448486328, + "learning_rate": 2.2152155424003952e-07, + "loss": 0.1013, + "num_input_tokens_seen": 9838296, + "step": 36695 + }, + { + "epoch": 9.619921363040628, + "grad_norm": 1.4588675498962402, + "learning_rate": 2.2000495896244976e-07, + "loss": 0.3366, + "num_input_tokens_seen": 9839912, + "step": 36700 + }, + { + "epoch": 9.621231979030144, + "grad_norm": 0.41060948371887207, + "learning_rate": 2.1849355009500582e-07, + "loss": 0.1301, + "num_input_tokens_seen": 9841096, + "step": 36705 + }, + { + "epoch": 9.62254259501966, + "grad_norm": 0.4984363317489624, + "learning_rate": 2.1698732795404074e-07, + "loss": 0.0969, + "num_input_tokens_seen": 9842600, + "step": 36710 + }, + { + "epoch": 9.623853211009175, + "grad_norm": 0.32649704813957214, + "learning_rate": 2.1548629285480792e-07, + "loss": 0.0826, + "num_input_tokens_seen": 9844456, + "step": 36715 + }, + { + "epoch": 9.625163826998689, + "grad_norm": 0.9191837310791016, + "learning_rate": 2.1399044511146993e-07, + "loss": 0.0845, + "num_input_tokens_seen": 9845944, + "step": 36720 + }, + { + "epoch": 9.626474442988204, + "grad_norm": 1.8267536163330078, + "learning_rate": 2.124997850371041e-07, + "loss": 0.1206, + "num_input_tokens_seen": 9847400, + "step": 36725 + }, + { + "epoch": 9.62778505897772, + "grad_norm": 1.0752177238464355, + "learning_rate": 2.1101431294370533e-07, + "loss": 0.1642, + "num_input_tokens_seen": 9848440, + "step": 36730 + }, + { + "epoch": 9.629095674967235, + "grad_norm": 0.310573548078537, + "learning_rate": 2.0953402914217768e-07, + "loss": 0.1629, + "num_input_tokens_seen": 9849816, + "step": 36735 + }, + { + "epoch": 9.63040629095675, + "grad_norm": 2.0434818267822266, + "learning_rate": 2.0805893394234e-07, + "loss": 0.1517, + "num_input_tokens_seen": 9851304, + "step": 36740 + }, + { + "epoch": 9.631716906946265, + "grad_norm": 1.2509905099868774, + "learning_rate": 2.065890276529342e-07, + "loss": 0.1043, + "num_input_tokens_seen": 9852568, + "step": 36745 + }, + { + "epoch": 9.63302752293578, + "grad_norm": 2.9210000038146973, + "learning_rate": 2.051243105816031e-07, + "loss": 0.1554, + "num_input_tokens_seen": 9854056, + "step": 36750 + }, + { + "epoch": 9.634338138925294, + "grad_norm": 0.600145161151886, + "learning_rate": 2.0366478303491255e-07, + "loss": 0.082, + "num_input_tokens_seen": 9855240, + "step": 36755 + }, + { + "epoch": 9.63564875491481, + "grad_norm": 2.165393114089966, + "learning_rate": 2.0221044531834043e-07, + "loss": 0.1635, + "num_input_tokens_seen": 9856760, + "step": 36760 + }, + { + "epoch": 9.636959370904325, + "grad_norm": 0.7077255249023438, + "learning_rate": 2.0076129773627105e-07, + "loss": 0.1282, + "num_input_tokens_seen": 9858360, + "step": 36765 + }, + { + "epoch": 9.63826998689384, + "grad_norm": 1.8760507106781006, + "learning_rate": 1.9931734059201457e-07, + "loss": 0.1528, + "num_input_tokens_seen": 9859624, + "step": 36770 + }, + { + "epoch": 9.639580602883354, + "grad_norm": 1.1270275115966797, + "learning_rate": 1.9787857418778478e-07, + "loss": 0.0667, + "num_input_tokens_seen": 9860856, + "step": 36775 + }, + { + "epoch": 9.64089121887287, + "grad_norm": 1.5560888051986694, + "learning_rate": 1.9644499882471578e-07, + "loss": 0.1241, + "num_input_tokens_seen": 9862296, + "step": 36780 + }, + { + "epoch": 9.642201834862385, + "grad_norm": 1.3218430280685425, + "learning_rate": 1.9501661480285095e-07, + "loss": 0.1271, + "num_input_tokens_seen": 9863784, + "step": 36785 + }, + { + "epoch": 9.643512450851901, + "grad_norm": 2.0590150356292725, + "learning_rate": 1.935934224211483e-07, + "loss": 0.1759, + "num_input_tokens_seen": 9865288, + "step": 36790 + }, + { + "epoch": 9.644823066841415, + "grad_norm": 3.194399118423462, + "learning_rate": 1.9217542197748074e-07, + "loss": 0.0839, + "num_input_tokens_seen": 9866504, + "step": 36795 + }, + { + "epoch": 9.64613368283093, + "grad_norm": 2.250671625137329, + "learning_rate": 1.9076261376863303e-07, + "loss": 0.1614, + "num_input_tokens_seen": 9867480, + "step": 36800 + }, + { + "epoch": 9.647444298820446, + "grad_norm": 1.5746763944625854, + "learning_rate": 1.8935499809030476e-07, + "loss": 0.1588, + "num_input_tokens_seen": 9869256, + "step": 36805 + }, + { + "epoch": 9.648754914809961, + "grad_norm": 0.7225951552391052, + "learning_rate": 1.8795257523710197e-07, + "loss": 0.1095, + "num_input_tokens_seen": 9871160, + "step": 36810 + }, + { + "epoch": 9.650065530799475, + "grad_norm": 1.4561117887496948, + "learning_rate": 1.8655534550255648e-07, + "loss": 0.1181, + "num_input_tokens_seen": 9872504, + "step": 36815 + }, + { + "epoch": 9.65137614678899, + "grad_norm": 2.6427690982818604, + "learning_rate": 1.8516330917910108e-07, + "loss": 0.0975, + "num_input_tokens_seen": 9873928, + "step": 36820 + }, + { + "epoch": 9.652686762778506, + "grad_norm": 5.263351917266846, + "learning_rate": 1.8377646655808877e-07, + "loss": 0.1904, + "num_input_tokens_seen": 9875368, + "step": 36825 + }, + { + "epoch": 9.653997378768022, + "grad_norm": 1.5760060548782349, + "learning_rate": 1.8239481792978464e-07, + "loss": 0.1047, + "num_input_tokens_seen": 9876584, + "step": 36830 + }, + { + "epoch": 9.655307994757536, + "grad_norm": 1.3111428022384644, + "learning_rate": 1.8101836358336287e-07, + "loss": 0.1226, + "num_input_tokens_seen": 9878584, + "step": 36835 + }, + { + "epoch": 9.656618610747051, + "grad_norm": 2.7541184425354004, + "learning_rate": 1.7964710380691251e-07, + "loss": 0.1224, + "num_input_tokens_seen": 9879432, + "step": 36840 + }, + { + "epoch": 9.657929226736567, + "grad_norm": 13.119538307189941, + "learning_rate": 1.782810388874373e-07, + "loss": 0.1134, + "num_input_tokens_seen": 9880744, + "step": 36845 + }, + { + "epoch": 9.65923984272608, + "grad_norm": 3.506650924682617, + "learning_rate": 1.7692016911085295e-07, + "loss": 0.1709, + "num_input_tokens_seen": 9882280, + "step": 36850 + }, + { + "epoch": 9.660550458715596, + "grad_norm": 1.233153223991394, + "learning_rate": 1.7556449476198445e-07, + "loss": 0.2123, + "num_input_tokens_seen": 9883816, + "step": 36855 + }, + { + "epoch": 9.661861074705111, + "grad_norm": 5.050828456878662, + "learning_rate": 1.7421401612457423e-07, + "loss": 0.1901, + "num_input_tokens_seen": 9885512, + "step": 36860 + }, + { + "epoch": 9.663171690694627, + "grad_norm": 1.757888674736023, + "learning_rate": 1.7286873348127676e-07, + "loss": 0.085, + "num_input_tokens_seen": 9886600, + "step": 36865 + }, + { + "epoch": 9.66448230668414, + "grad_norm": 2.4150454998016357, + "learning_rate": 1.715286471136529e-07, + "loss": 0.1106, + "num_input_tokens_seen": 9887592, + "step": 36870 + }, + { + "epoch": 9.665792922673656, + "grad_norm": 2.592050552368164, + "learning_rate": 1.7019375730218667e-07, + "loss": 0.1021, + "num_input_tokens_seen": 9888504, + "step": 36875 + }, + { + "epoch": 9.667103538663172, + "grad_norm": 2.1064107418060303, + "learning_rate": 1.688640643262629e-07, + "loss": 0.164, + "num_input_tokens_seen": 9889688, + "step": 36880 + }, + { + "epoch": 9.668414154652687, + "grad_norm": 1.0959450006484985, + "learning_rate": 1.6753956846418395e-07, + "loss": 0.2311, + "num_input_tokens_seen": 9891304, + "step": 36885 + }, + { + "epoch": 9.669724770642201, + "grad_norm": 2.274139881134033, + "learning_rate": 1.6622026999316697e-07, + "loss": 0.104, + "num_input_tokens_seen": 9893064, + "step": 36890 + }, + { + "epoch": 9.671035386631717, + "grad_norm": 2.023510456085205, + "learning_rate": 1.6490616918933554e-07, + "loss": 0.1408, + "num_input_tokens_seen": 9894680, + "step": 36895 + }, + { + "epoch": 9.672346002621232, + "grad_norm": 1.2414119243621826, + "learning_rate": 1.6359726632773353e-07, + "loss": 0.1382, + "num_input_tokens_seen": 9896056, + "step": 36900 + }, + { + "epoch": 9.673656618610748, + "grad_norm": 2.7286605834960938, + "learning_rate": 1.6229356168231125e-07, + "loss": 0.1903, + "num_input_tokens_seen": 9897448, + "step": 36905 + }, + { + "epoch": 9.674967234600262, + "grad_norm": 1.695290446281433, + "learning_rate": 1.609950555259282e-07, + "loss": 0.1601, + "num_input_tokens_seen": 9899128, + "step": 36910 + }, + { + "epoch": 9.676277850589777, + "grad_norm": 1.523741364479065, + "learning_rate": 1.5970174813036144e-07, + "loss": 0.1448, + "num_input_tokens_seen": 9900552, + "step": 36915 + }, + { + "epoch": 9.677588466579293, + "grad_norm": 1.5267279148101807, + "learning_rate": 1.5841363976629998e-07, + "loss": 0.1916, + "num_input_tokens_seen": 9901784, + "step": 36920 + }, + { + "epoch": 9.678899082568808, + "grad_norm": 1.3501163721084595, + "learning_rate": 1.5713073070333927e-07, + "loss": 0.1158, + "num_input_tokens_seen": 9902968, + "step": 36925 + }, + { + "epoch": 9.680209698558322, + "grad_norm": 2.7373921871185303, + "learning_rate": 1.5585302120998956e-07, + "loss": 0.1522, + "num_input_tokens_seen": 9904472, + "step": 36930 + }, + { + "epoch": 9.681520314547837, + "grad_norm": 1.8205314874649048, + "learning_rate": 1.545805115536786e-07, + "loss": 0.1691, + "num_input_tokens_seen": 9905768, + "step": 36935 + }, + { + "epoch": 9.682830930537353, + "grad_norm": 1.4884980916976929, + "learning_rate": 1.5331320200073497e-07, + "loss": 0.1111, + "num_input_tokens_seen": 9907912, + "step": 36940 + }, + { + "epoch": 9.684141546526867, + "grad_norm": 1.6667025089263916, + "learning_rate": 1.520510928164076e-07, + "loss": 0.0536, + "num_input_tokens_seen": 9908984, + "step": 36945 + }, + { + "epoch": 9.685452162516382, + "grad_norm": 4.798120021820068, + "learning_rate": 1.5079418426485193e-07, + "loss": 0.2044, + "num_input_tokens_seen": 9910216, + "step": 36950 + }, + { + "epoch": 9.686762778505898, + "grad_norm": 3.192345142364502, + "learning_rate": 1.49542476609138e-07, + "loss": 0.1604, + "num_input_tokens_seen": 9911544, + "step": 36955 + }, + { + "epoch": 9.688073394495413, + "grad_norm": 1.639568567276001, + "learning_rate": 1.482959701112452e-07, + "loss": 0.1311, + "num_input_tokens_seen": 9913352, + "step": 36960 + }, + { + "epoch": 9.689384010484927, + "grad_norm": 0.5911386609077454, + "learning_rate": 1.47054665032062e-07, + "loss": 0.1137, + "num_input_tokens_seen": 9914376, + "step": 36965 + }, + { + "epoch": 9.690694626474443, + "grad_norm": 1.8213390111923218, + "learning_rate": 1.4581856163140008e-07, + "loss": 0.1273, + "num_input_tokens_seen": 9915448, + "step": 36970 + }, + { + "epoch": 9.692005242463958, + "grad_norm": 1.9171499013900757, + "learning_rate": 1.4458766016796632e-07, + "loss": 0.1952, + "num_input_tokens_seen": 9916936, + "step": 36975 + }, + { + "epoch": 9.693315858453474, + "grad_norm": 4.98536491394043, + "learning_rate": 1.4336196089938802e-07, + "loss": 0.1393, + "num_input_tokens_seen": 9918200, + "step": 36980 + }, + { + "epoch": 9.694626474442988, + "grad_norm": 1.3941580057144165, + "learning_rate": 1.4214146408220163e-07, + "loss": 0.1943, + "num_input_tokens_seen": 9920120, + "step": 36985 + }, + { + "epoch": 9.695937090432503, + "grad_norm": 1.8837436437606812, + "learning_rate": 1.4092616997185837e-07, + "loss": 0.1336, + "num_input_tokens_seen": 9921464, + "step": 36990 + }, + { + "epoch": 9.697247706422019, + "grad_norm": 1.6719090938568115, + "learning_rate": 1.397160788227131e-07, + "loss": 0.2762, + "num_input_tokens_seen": 9923128, + "step": 36995 + }, + { + "epoch": 9.698558322411534, + "grad_norm": 3.2729756832122803, + "learning_rate": 1.385111908880382e-07, + "loss": 0.1702, + "num_input_tokens_seen": 9924408, + "step": 37000 + }, + { + "epoch": 9.699868938401048, + "grad_norm": 1.5310310125350952, + "learning_rate": 1.3731150642001255e-07, + "loss": 0.2028, + "num_input_tokens_seen": 9925912, + "step": 37005 + }, + { + "epoch": 9.701179554390563, + "grad_norm": 2.7350504398345947, + "learning_rate": 1.3611702566972694e-07, + "loss": 0.1353, + "num_input_tokens_seen": 9927000, + "step": 37010 + }, + { + "epoch": 9.702490170380079, + "grad_norm": 2.6938652992248535, + "learning_rate": 1.3492774888718974e-07, + "loss": 0.1407, + "num_input_tokens_seen": 9927992, + "step": 37015 + }, + { + "epoch": 9.703800786369595, + "grad_norm": 1.2903766632080078, + "learning_rate": 1.3374367632131014e-07, + "loss": 0.2649, + "num_input_tokens_seen": 9929448, + "step": 37020 + }, + { + "epoch": 9.705111402359108, + "grad_norm": 5.96091890335083, + "learning_rate": 1.3256480821991213e-07, + "loss": 0.1376, + "num_input_tokens_seen": 9930808, + "step": 37025 + }, + { + "epoch": 9.706422018348624, + "grad_norm": 1.4445598125457764, + "learning_rate": 1.3139114482973448e-07, + "loss": 0.1696, + "num_input_tokens_seen": 9932232, + "step": 37030 + }, + { + "epoch": 9.70773263433814, + "grad_norm": 1.1365107297897339, + "learning_rate": 1.3022268639641956e-07, + "loss": 0.2084, + "num_input_tokens_seen": 9933656, + "step": 37035 + }, + { + "epoch": 9.709043250327653, + "grad_norm": 2.4845969676971436, + "learning_rate": 1.290594331645245e-07, + "loss": 0.1302, + "num_input_tokens_seen": 9935368, + "step": 37040 + }, + { + "epoch": 9.710353866317169, + "grad_norm": 3.677619218826294, + "learning_rate": 1.279013853775185e-07, + "loss": 0.0587, + "num_input_tokens_seen": 9936376, + "step": 37045 + }, + { + "epoch": 9.711664482306684, + "grad_norm": 1.1251236200332642, + "learning_rate": 1.267485432777743e-07, + "loss": 0.1608, + "num_input_tokens_seen": 9937464, + "step": 37050 + }, + { + "epoch": 9.7129750982962, + "grad_norm": 0.3087571859359741, + "learning_rate": 1.2560090710658223e-07, + "loss": 0.1002, + "num_input_tokens_seen": 9938696, + "step": 37055 + }, + { + "epoch": 9.714285714285714, + "grad_norm": 1.0748422145843506, + "learning_rate": 1.2445847710414183e-07, + "loss": 0.0309, + "num_input_tokens_seen": 9939656, + "step": 37060 + }, + { + "epoch": 9.715596330275229, + "grad_norm": 1.7080283164978027, + "learning_rate": 1.233212535095618e-07, + "loss": 0.1133, + "num_input_tokens_seen": 9940888, + "step": 37065 + }, + { + "epoch": 9.716906946264745, + "grad_norm": 2.0916569232940674, + "learning_rate": 1.2218923656085735e-07, + "loss": 0.0987, + "num_input_tokens_seen": 9942024, + "step": 37070 + }, + { + "epoch": 9.71821756225426, + "grad_norm": 0.7268698811531067, + "learning_rate": 1.2106242649496112e-07, + "loss": 0.1676, + "num_input_tokens_seen": 9943336, + "step": 37075 + }, + { + "epoch": 9.719528178243774, + "grad_norm": 1.2664576768875122, + "learning_rate": 1.199408235477123e-07, + "loss": 0.3596, + "num_input_tokens_seen": 9945096, + "step": 37080 + }, + { + "epoch": 9.72083879423329, + "grad_norm": 1.8947782516479492, + "learning_rate": 1.188244279538564e-07, + "loss": 0.1564, + "num_input_tokens_seen": 9946344, + "step": 37085 + }, + { + "epoch": 9.722149410222805, + "grad_norm": 2.291903257369995, + "learning_rate": 1.1771323994705929e-07, + "loss": 0.0906, + "num_input_tokens_seen": 9947688, + "step": 37090 + }, + { + "epoch": 9.72346002621232, + "grad_norm": 1.2928744554519653, + "learning_rate": 1.1660725975988773e-07, + "loss": 0.1409, + "num_input_tokens_seen": 9949816, + "step": 37095 + }, + { + "epoch": 9.724770642201834, + "grad_norm": 1.9110710620880127, + "learning_rate": 1.1550648762382044e-07, + "loss": 0.2164, + "num_input_tokens_seen": 9951112, + "step": 37100 + }, + { + "epoch": 9.72608125819135, + "grad_norm": 2.786468029022217, + "learning_rate": 1.144109237692509e-07, + "loss": 0.1131, + "num_input_tokens_seen": 9952312, + "step": 37105 + }, + { + "epoch": 9.727391874180865, + "grad_norm": 4.376785755157471, + "learning_rate": 1.1332056842547344e-07, + "loss": 0.1201, + "num_input_tokens_seen": 9953448, + "step": 37110 + }, + { + "epoch": 9.728702490170381, + "grad_norm": 2.168811798095703, + "learning_rate": 1.1223542182070002e-07, + "loss": 0.1869, + "num_input_tokens_seen": 9954936, + "step": 37115 + }, + { + "epoch": 9.730013106159895, + "grad_norm": 1.3404101133346558, + "learning_rate": 1.1115548418205168e-07, + "loss": 0.2003, + "num_input_tokens_seen": 9956216, + "step": 37120 + }, + { + "epoch": 9.73132372214941, + "grad_norm": 4.405375957489014, + "learning_rate": 1.1008075573555599e-07, + "loss": 0.1544, + "num_input_tokens_seen": 9957416, + "step": 37125 + }, + { + "epoch": 9.732634338138926, + "grad_norm": 3.719705820083618, + "learning_rate": 1.0901123670614965e-07, + "loss": 0.0887, + "num_input_tokens_seen": 9958504, + "step": 37130 + }, + { + "epoch": 9.73394495412844, + "grad_norm": 2.616227865219116, + "learning_rate": 1.0794692731768419e-07, + "loss": 0.1886, + "num_input_tokens_seen": 9959784, + "step": 37135 + }, + { + "epoch": 9.735255570117955, + "grad_norm": 1.8563119173049927, + "learning_rate": 1.0688782779291473e-07, + "loss": 0.2485, + "num_input_tokens_seen": 9961576, + "step": 37140 + }, + { + "epoch": 9.73656618610747, + "grad_norm": 1.5120245218276978, + "learning_rate": 1.0583393835351396e-07, + "loss": 0.1229, + "num_input_tokens_seen": 9963032, + "step": 37145 + }, + { + "epoch": 9.737876802096986, + "grad_norm": 0.9208672642707825, + "learning_rate": 1.0478525922005545e-07, + "loss": 0.0875, + "num_input_tokens_seen": 9964056, + "step": 37150 + }, + { + "epoch": 9.7391874180865, + "grad_norm": 2.290156364440918, + "learning_rate": 1.0374179061202749e-07, + "loss": 0.065, + "num_input_tokens_seen": 9965368, + "step": 37155 + }, + { + "epoch": 9.740498034076015, + "grad_norm": 2.8543388843536377, + "learning_rate": 1.0270353274782207e-07, + "loss": 0.1973, + "num_input_tokens_seen": 9966680, + "step": 37160 + }, + { + "epoch": 9.741808650065531, + "grad_norm": 2.934638023376465, + "learning_rate": 1.0167048584475147e-07, + "loss": 0.1115, + "num_input_tokens_seen": 9967864, + "step": 37165 + }, + { + "epoch": 9.743119266055047, + "grad_norm": 1.7542774677276611, + "learning_rate": 1.006426501190233e-07, + "loss": 0.1481, + "num_input_tokens_seen": 9969448, + "step": 37170 + }, + { + "epoch": 9.74442988204456, + "grad_norm": 3.132503032684326, + "learning_rate": 9.962002578576823e-08, + "loss": 0.1174, + "num_input_tokens_seen": 9970712, + "step": 37175 + }, + { + "epoch": 9.745740498034076, + "grad_norm": 2.034196376800537, + "learning_rate": 9.860261305901785e-08, + "loss": 0.1594, + "num_input_tokens_seen": 9971736, + "step": 37180 + }, + { + "epoch": 9.747051114023591, + "grad_norm": 4.933801174163818, + "learning_rate": 9.759041215171295e-08, + "loss": 0.1186, + "num_input_tokens_seen": 9972872, + "step": 37185 + }, + { + "epoch": 9.748361730013107, + "grad_norm": 2.4445667266845703, + "learning_rate": 9.658342327570902e-08, + "loss": 0.2553, + "num_input_tokens_seen": 9973784, + "step": 37190 + }, + { + "epoch": 9.74967234600262, + "grad_norm": 1.2621010541915894, + "learning_rate": 9.55816466417625e-08, + "loss": 0.0937, + "num_input_tokens_seen": 9974840, + "step": 37195 + }, + { + "epoch": 9.750982961992136, + "grad_norm": 5.375865459442139, + "learning_rate": 9.458508245954456e-08, + "loss": 0.1639, + "num_input_tokens_seen": 9976088, + "step": 37200 + }, + { + "epoch": 9.752293577981652, + "grad_norm": 2.170334577560425, + "learning_rate": 9.359373093763835e-08, + "loss": 0.1298, + "num_input_tokens_seen": 9977432, + "step": 37205 + }, + { + "epoch": 9.753604193971167, + "grad_norm": 3.204257011413574, + "learning_rate": 9.260759228352789e-08, + "loss": 0.1769, + "num_input_tokens_seen": 9978392, + "step": 37210 + }, + { + "epoch": 9.754914809960681, + "grad_norm": 3.011526584625244, + "learning_rate": 9.162666670361198e-08, + "loss": 0.2187, + "num_input_tokens_seen": 9980024, + "step": 37215 + }, + { + "epoch": 9.756225425950197, + "grad_norm": 3.6756575107574463, + "learning_rate": 9.065095440319582e-08, + "loss": 0.1725, + "num_input_tokens_seen": 9981416, + "step": 37220 + }, + { + "epoch": 9.757536041939712, + "grad_norm": 1.906554102897644, + "learning_rate": 8.968045558649663e-08, + "loss": 0.0916, + "num_input_tokens_seen": 9982664, + "step": 37225 + }, + { + "epoch": 9.758846657929226, + "grad_norm": 2.1900556087493896, + "learning_rate": 8.871517045663524e-08, + "loss": 0.096, + "num_input_tokens_seen": 9984040, + "step": 37230 + }, + { + "epoch": 9.760157273918741, + "grad_norm": 2.6185436248779297, + "learning_rate": 8.775509921564728e-08, + "loss": 0.1811, + "num_input_tokens_seen": 9985400, + "step": 37235 + }, + { + "epoch": 9.761467889908257, + "grad_norm": 4.3953423500061035, + "learning_rate": 8.680024206446924e-08, + "loss": 0.26, + "num_input_tokens_seen": 9986584, + "step": 37240 + }, + { + "epoch": 9.762778505897773, + "grad_norm": 2.907273530960083, + "learning_rate": 8.585059920295512e-08, + "loss": 0.2173, + "num_input_tokens_seen": 9988088, + "step": 37245 + }, + { + "epoch": 9.764089121887286, + "grad_norm": 3.5819098949432373, + "learning_rate": 8.490617082985986e-08, + "loss": 0.2054, + "num_input_tokens_seen": 9989192, + "step": 37250 + }, + { + "epoch": 9.765399737876802, + "grad_norm": 2.051544427871704, + "learning_rate": 8.396695714285585e-08, + "loss": 0.1346, + "num_input_tokens_seen": 9990392, + "step": 37255 + }, + { + "epoch": 9.766710353866317, + "grad_norm": 1.7831050157546997, + "learning_rate": 8.303295833851365e-08, + "loss": 0.1504, + "num_input_tokens_seen": 9991608, + "step": 37260 + }, + { + "epoch": 9.768020969855833, + "grad_norm": 1.1401981115341187, + "learning_rate": 8.210417461232412e-08, + "loss": 0.1396, + "num_input_tokens_seen": 9992968, + "step": 37265 + }, + { + "epoch": 9.769331585845347, + "grad_norm": 1.186316967010498, + "learning_rate": 8.11806061586734e-08, + "loss": 0.1273, + "num_input_tokens_seen": 9994280, + "step": 37270 + }, + { + "epoch": 9.770642201834862, + "grad_norm": 1.8784019947052002, + "learning_rate": 8.026225317086522e-08, + "loss": 0.1413, + "num_input_tokens_seen": 9995736, + "step": 37275 + }, + { + "epoch": 9.771952817824378, + "grad_norm": 3.3035802841186523, + "learning_rate": 7.934911584110971e-08, + "loss": 0.1667, + "num_input_tokens_seen": 9997064, + "step": 37280 + }, + { + "epoch": 9.773263433813893, + "grad_norm": 1.219872236251831, + "learning_rate": 7.844119436052622e-08, + "loss": 0.1005, + "num_input_tokens_seen": 9998072, + "step": 37285 + }, + { + "epoch": 9.774574049803407, + "grad_norm": 3.5840611457824707, + "learning_rate": 7.753848891913772e-08, + "loss": 0.1651, + "num_input_tokens_seen": 9999416, + "step": 37290 + }, + { + "epoch": 9.775884665792923, + "grad_norm": 2.470651388168335, + "learning_rate": 7.664099970588201e-08, + "loss": 0.1225, + "num_input_tokens_seen": 10000648, + "step": 37295 + }, + { + "epoch": 9.777195281782438, + "grad_norm": 0.6649083495140076, + "learning_rate": 7.57487269085977e-08, + "loss": 0.0339, + "num_input_tokens_seen": 10001544, + "step": 37300 + }, + { + "epoch": 9.778505897771954, + "grad_norm": 2.591550350189209, + "learning_rate": 7.486167071404371e-08, + "loss": 0.0857, + "num_input_tokens_seen": 10002648, + "step": 37305 + }, + { + "epoch": 9.779816513761467, + "grad_norm": 4.933570861816406, + "learning_rate": 7.397983130787156e-08, + "loss": 0.1473, + "num_input_tokens_seen": 10004024, + "step": 37310 + }, + { + "epoch": 9.781127129750983, + "grad_norm": 7.408143997192383, + "learning_rate": 7.310320887465305e-08, + "loss": 0.1754, + "num_input_tokens_seen": 10005288, + "step": 37315 + }, + { + "epoch": 9.782437745740499, + "grad_norm": 1.6199771165847778, + "learning_rate": 7.223180359786086e-08, + "loss": 0.1704, + "num_input_tokens_seen": 10006600, + "step": 37320 + }, + { + "epoch": 9.783748361730012, + "grad_norm": 1.9403389692306519, + "learning_rate": 7.136561565988242e-08, + "loss": 0.1791, + "num_input_tokens_seen": 10008328, + "step": 37325 + }, + { + "epoch": 9.785058977719528, + "grad_norm": 1.3752652406692505, + "learning_rate": 7.050464524200607e-08, + "loss": 0.1542, + "num_input_tokens_seen": 10009528, + "step": 37330 + }, + { + "epoch": 9.786369593709043, + "grad_norm": 1.15230393409729, + "learning_rate": 6.964889252443485e-08, + "loss": 0.1194, + "num_input_tokens_seen": 10010920, + "step": 37335 + }, + { + "epoch": 9.787680209698559, + "grad_norm": 2.7568769454956055, + "learning_rate": 6.879835768627274e-08, + "loss": 0.2579, + "num_input_tokens_seen": 10012584, + "step": 37340 + }, + { + "epoch": 9.788990825688073, + "grad_norm": 3.281132936477661, + "learning_rate": 6.795304090553567e-08, + "loss": 0.1656, + "num_input_tokens_seen": 10013864, + "step": 37345 + }, + { + "epoch": 9.790301441677588, + "grad_norm": 2.3474326133728027, + "learning_rate": 6.711294235914877e-08, + "loss": 0.2039, + "num_input_tokens_seen": 10016008, + "step": 37350 + }, + { + "epoch": 9.791612057667104, + "grad_norm": 1.4487791061401367, + "learning_rate": 6.627806222294086e-08, + "loss": 0.1005, + "num_input_tokens_seen": 10017176, + "step": 37355 + }, + { + "epoch": 9.79292267365662, + "grad_norm": 1.942677617073059, + "learning_rate": 6.544840067165548e-08, + "loss": 0.1054, + "num_input_tokens_seen": 10018584, + "step": 37360 + }, + { + "epoch": 9.794233289646133, + "grad_norm": 1.0553754568099976, + "learning_rate": 6.462395787893427e-08, + "loss": 0.1791, + "num_input_tokens_seen": 10020392, + "step": 37365 + }, + { + "epoch": 9.795543905635649, + "grad_norm": 1.0022846460342407, + "learning_rate": 6.380473401733366e-08, + "loss": 0.1947, + "num_input_tokens_seen": 10022424, + "step": 37370 + }, + { + "epoch": 9.796854521625164, + "grad_norm": 1.2649624347686768, + "learning_rate": 6.299072925831373e-08, + "loss": 0.1129, + "num_input_tokens_seen": 10024296, + "step": 37375 + }, + { + "epoch": 9.79816513761468, + "grad_norm": 5.368220329284668, + "learning_rate": 6.218194377224928e-08, + "loss": 0.1901, + "num_input_tokens_seen": 10025304, + "step": 37380 + }, + { + "epoch": 9.799475753604193, + "grad_norm": 3.4776113033294678, + "learning_rate": 6.137837772841326e-08, + "loss": 0.0631, + "num_input_tokens_seen": 10026344, + "step": 37385 + }, + { + "epoch": 9.800786369593709, + "grad_norm": 3.898798704147339, + "learning_rate": 6.058003129499334e-08, + "loss": 0.1662, + "num_input_tokens_seen": 10027464, + "step": 37390 + }, + { + "epoch": 9.802096985583225, + "grad_norm": 3.1518642902374268, + "learning_rate": 5.978690463908088e-08, + "loss": 0.1315, + "num_input_tokens_seen": 10028680, + "step": 37395 + }, + { + "epoch": 9.80340760157274, + "grad_norm": 3.157335042953491, + "learning_rate": 5.8998997926676405e-08, + "loss": 0.1931, + "num_input_tokens_seen": 10029928, + "step": 37400 + }, + { + "epoch": 9.804718217562254, + "grad_norm": 0.19510096311569214, + "learning_rate": 5.821631132268412e-08, + "loss": 0.1339, + "num_input_tokens_seen": 10031208, + "step": 37405 + }, + { + "epoch": 9.80602883355177, + "grad_norm": 3.117640256881714, + "learning_rate": 5.743884499092578e-08, + "loss": 0.1047, + "num_input_tokens_seen": 10032136, + "step": 37410 + }, + { + "epoch": 9.807339449541285, + "grad_norm": 2.6682136058807373, + "learning_rate": 5.6666599094115646e-08, + "loss": 0.1277, + "num_input_tokens_seen": 10033176, + "step": 37415 + }, + { + "epoch": 9.808650065530799, + "grad_norm": 0.16237403452396393, + "learning_rate": 5.58995737938911e-08, + "loss": 0.1038, + "num_input_tokens_seen": 10034280, + "step": 37420 + }, + { + "epoch": 9.809960681520314, + "grad_norm": 0.8122776746749878, + "learning_rate": 5.513776925078207e-08, + "loss": 0.1245, + "num_input_tokens_seen": 10035960, + "step": 37425 + }, + { + "epoch": 9.81127129750983, + "grad_norm": 3.8848135471343994, + "learning_rate": 5.4381185624238776e-08, + "loss": 0.07, + "num_input_tokens_seen": 10036984, + "step": 37430 + }, + { + "epoch": 9.812581913499345, + "grad_norm": 3.1363401412963867, + "learning_rate": 5.362982307261233e-08, + "loss": 0.1575, + "num_input_tokens_seen": 10038008, + "step": 37435 + }, + { + "epoch": 9.813892529488859, + "grad_norm": 1.120895266532898, + "learning_rate": 5.2883681753157497e-08, + "loss": 0.1566, + "num_input_tokens_seen": 10039112, + "step": 37440 + }, + { + "epoch": 9.815203145478375, + "grad_norm": 2.3299601078033447, + "learning_rate": 5.214276182204381e-08, + "loss": 0.0962, + "num_input_tokens_seen": 10040216, + "step": 37445 + }, + { + "epoch": 9.81651376146789, + "grad_norm": 1.8542790412902832, + "learning_rate": 5.140706343434165e-08, + "loss": 0.1576, + "num_input_tokens_seen": 10041528, + "step": 37450 + }, + { + "epoch": 9.817824377457406, + "grad_norm": 5.2857866287231445, + "learning_rate": 5.0676586744036194e-08, + "loss": 0.1097, + "num_input_tokens_seen": 10042600, + "step": 37455 + }, + { + "epoch": 9.81913499344692, + "grad_norm": 2.179503917694092, + "learning_rate": 4.9951331904007915e-08, + "loss": 0.1815, + "num_input_tokens_seen": 10043960, + "step": 37460 + }, + { + "epoch": 9.820445609436435, + "grad_norm": 4.375009059906006, + "learning_rate": 4.923129906606039e-08, + "loss": 0.1454, + "num_input_tokens_seen": 10045080, + "step": 37465 + }, + { + "epoch": 9.82175622542595, + "grad_norm": 0.5479637384414673, + "learning_rate": 4.851648838088696e-08, + "loss": 0.1158, + "num_input_tokens_seen": 10046440, + "step": 37470 + }, + { + "epoch": 9.823066841415466, + "grad_norm": 2.874971866607666, + "learning_rate": 4.7806899998101283e-08, + "loss": 0.1061, + "num_input_tokens_seen": 10047416, + "step": 37475 + }, + { + "epoch": 9.82437745740498, + "grad_norm": 1.001488208770752, + "learning_rate": 4.71025340662179e-08, + "loss": 0.1268, + "num_input_tokens_seen": 10048648, + "step": 37480 + }, + { + "epoch": 9.825688073394495, + "grad_norm": 1.8957276344299316, + "learning_rate": 4.6403390732654985e-08, + "loss": 0.1763, + "num_input_tokens_seen": 10050008, + "step": 37485 + }, + { + "epoch": 9.82699868938401, + "grad_norm": 3.616157054901123, + "learning_rate": 4.570947014374827e-08, + "loss": 0.1758, + "num_input_tokens_seen": 10051752, + "step": 37490 + }, + { + "epoch": 9.828309305373526, + "grad_norm": 2.6437811851501465, + "learning_rate": 4.502077244473435e-08, + "loss": 0.1611, + "num_input_tokens_seen": 10053032, + "step": 37495 + }, + { + "epoch": 9.82961992136304, + "grad_norm": 1.4273492097854614, + "learning_rate": 4.4337297779750705e-08, + "loss": 0.0778, + "num_input_tokens_seen": 10054104, + "step": 37500 + }, + { + "epoch": 9.830930537352556, + "grad_norm": 5.230043411254883, + "learning_rate": 4.365904629185236e-08, + "loss": 0.0905, + "num_input_tokens_seen": 10055288, + "step": 37505 + }, + { + "epoch": 9.832241153342071, + "grad_norm": 1.531302571296692, + "learning_rate": 4.298601812299241e-08, + "loss": 0.1889, + "num_input_tokens_seen": 10056808, + "step": 37510 + }, + { + "epoch": 9.833551769331585, + "grad_norm": 2.947301149368286, + "learning_rate": 4.2318213414038745e-08, + "loss": 0.1385, + "num_input_tokens_seen": 10057960, + "step": 37515 + }, + { + "epoch": 9.8348623853211, + "grad_norm": 1.334384799003601, + "learning_rate": 4.1655632304757334e-08, + "loss": 0.1577, + "num_input_tokens_seen": 10059544, + "step": 37520 + }, + { + "epoch": 9.836173001310616, + "grad_norm": 2.1640350818634033, + "learning_rate": 4.0998274933828905e-08, + "loss": 0.0989, + "num_input_tokens_seen": 10060680, + "step": 37525 + }, + { + "epoch": 9.837483617300132, + "grad_norm": 3.299464702606201, + "learning_rate": 4.034614143883508e-08, + "loss": 0.1692, + "num_input_tokens_seen": 10061848, + "step": 37530 + }, + { + "epoch": 9.838794233289645, + "grad_norm": 2.0391666889190674, + "learning_rate": 3.969923195626668e-08, + "loss": 0.1286, + "num_input_tokens_seen": 10063224, + "step": 37535 + }, + { + "epoch": 9.840104849279161, + "grad_norm": 3.9037232398986816, + "learning_rate": 3.9057546621520946e-08, + "loss": 0.1245, + "num_input_tokens_seen": 10064728, + "step": 37540 + }, + { + "epoch": 9.841415465268676, + "grad_norm": 1.7855064868927002, + "learning_rate": 3.842108556890156e-08, + "loss": 0.062, + "num_input_tokens_seen": 10065832, + "step": 37545 + }, + { + "epoch": 9.842726081258192, + "grad_norm": 1.6778062582015991, + "learning_rate": 3.778984893161863e-08, + "loss": 0.1085, + "num_input_tokens_seen": 10067112, + "step": 37550 + }, + { + "epoch": 9.844036697247706, + "grad_norm": 3.69474458694458, + "learning_rate": 3.7163836841791476e-08, + "loss": 0.2165, + "num_input_tokens_seen": 10068264, + "step": 37555 + }, + { + "epoch": 9.845347313237221, + "grad_norm": 2.2358226776123047, + "learning_rate": 3.654304943043752e-08, + "loss": 0.173, + "num_input_tokens_seen": 10069448, + "step": 37560 + }, + { + "epoch": 9.846657929226737, + "grad_norm": 11.434229850769043, + "learning_rate": 3.5927486827491696e-08, + "loss": 0.2022, + "num_input_tokens_seen": 10070776, + "step": 37565 + }, + { + "epoch": 9.847968545216252, + "grad_norm": 0.5594025254249573, + "learning_rate": 3.531714916178708e-08, + "loss": 0.0696, + "num_input_tokens_seen": 10071928, + "step": 37570 + }, + { + "epoch": 9.849279161205766, + "grad_norm": 1.976802110671997, + "learning_rate": 3.4712036561068693e-08, + "loss": 0.1198, + "num_input_tokens_seen": 10072888, + "step": 37575 + }, + { + "epoch": 9.850589777195282, + "grad_norm": 0.6158620715141296, + "learning_rate": 3.4112149151982466e-08, + "loss": 0.058, + "num_input_tokens_seen": 10074008, + "step": 37580 + }, + { + "epoch": 9.851900393184797, + "grad_norm": 2.674381971359253, + "learning_rate": 3.351748706008628e-08, + "loss": 0.1727, + "num_input_tokens_seen": 10075544, + "step": 37585 + }, + { + "epoch": 9.853211009174313, + "grad_norm": 2.3288674354553223, + "learning_rate": 3.292805040984171e-08, + "loss": 0.2012, + "num_input_tokens_seen": 10076920, + "step": 37590 + }, + { + "epoch": 9.854521625163827, + "grad_norm": 2.0432114601135254, + "learning_rate": 3.234383932461671e-08, + "loss": 0.1669, + "num_input_tokens_seen": 10078456, + "step": 37595 + }, + { + "epoch": 9.855832241153342, + "grad_norm": 0.7109007239341736, + "learning_rate": 3.176485392668571e-08, + "loss": 0.1817, + "num_input_tokens_seen": 10079896, + "step": 37600 + }, + { + "epoch": 9.857142857142858, + "grad_norm": 0.4009828567504883, + "learning_rate": 3.119109433722955e-08, + "loss": 0.1492, + "num_input_tokens_seen": 10081512, + "step": 37605 + }, + { + "epoch": 9.858453473132371, + "grad_norm": 3.016596555709839, + "learning_rate": 3.0622560676332734e-08, + "loss": 0.333, + "num_input_tokens_seen": 10083096, + "step": 37610 + }, + { + "epoch": 9.859764089121887, + "grad_norm": 1.5687628984451294, + "learning_rate": 3.005925306299173e-08, + "loss": 0.163, + "num_input_tokens_seen": 10084632, + "step": 37615 + }, + { + "epoch": 9.861074705111402, + "grad_norm": 2.3333308696746826, + "learning_rate": 2.9501171615103907e-08, + "loss": 0.1332, + "num_input_tokens_seen": 10086152, + "step": 37620 + }, + { + "epoch": 9.862385321100918, + "grad_norm": 2.391964912414551, + "learning_rate": 2.8948316449473044e-08, + "loss": 0.1895, + "num_input_tokens_seen": 10087864, + "step": 37625 + }, + { + "epoch": 9.863695937090432, + "grad_norm": 1.8834757804870605, + "learning_rate": 2.840068768181492e-08, + "loss": 0.186, + "num_input_tokens_seen": 10089288, + "step": 37630 + }, + { + "epoch": 9.865006553079947, + "grad_norm": 1.1241337060928345, + "learning_rate": 2.785828542674618e-08, + "loss": 0.1775, + "num_input_tokens_seen": 10090552, + "step": 37635 + }, + { + "epoch": 9.866317169069463, + "grad_norm": 1.4118906259536743, + "learning_rate": 2.7321109797787125e-08, + "loss": 0.1587, + "num_input_tokens_seen": 10091944, + "step": 37640 + }, + { + "epoch": 9.867627785058978, + "grad_norm": 2.1514885425567627, + "learning_rate": 2.6789160907372822e-08, + "loss": 0.1536, + "num_input_tokens_seen": 10093400, + "step": 37645 + }, + { + "epoch": 9.868938401048492, + "grad_norm": 1.028665542602539, + "learning_rate": 2.626243886683366e-08, + "loss": 0.1645, + "num_input_tokens_seen": 10094536, + "step": 37650 + }, + { + "epoch": 9.870249017038008, + "grad_norm": 2.426726818084717, + "learning_rate": 2.574094378641756e-08, + "loss": 0.1396, + "num_input_tokens_seen": 10095736, + "step": 37655 + }, + { + "epoch": 9.871559633027523, + "grad_norm": 4.21833610534668, + "learning_rate": 2.5224675775270544e-08, + "loss": 0.2736, + "num_input_tokens_seen": 10096888, + "step": 37660 + }, + { + "epoch": 9.872870249017039, + "grad_norm": 2.230989694595337, + "learning_rate": 2.4713634941442298e-08, + "loss": 0.0679, + "num_input_tokens_seen": 10098072, + "step": 37665 + }, + { + "epoch": 9.874180865006553, + "grad_norm": 1.814134955406189, + "learning_rate": 2.4207821391900032e-08, + "loss": 0.2254, + "num_input_tokens_seen": 10099832, + "step": 37670 + }, + { + "epoch": 9.875491480996068, + "grad_norm": 2.3690543174743652, + "learning_rate": 2.370723523250351e-08, + "loss": 0.0752, + "num_input_tokens_seen": 10101352, + "step": 37675 + }, + { + "epoch": 9.876802096985584, + "grad_norm": 1.595574140548706, + "learning_rate": 2.321187656802726e-08, + "loss": 0.1054, + "num_input_tokens_seen": 10102632, + "step": 37680 + }, + { + "epoch": 9.8781127129751, + "grad_norm": 0.4690393805503845, + "learning_rate": 2.272174550214945e-08, + "loss": 0.1462, + "num_input_tokens_seen": 10103688, + "step": 37685 + }, + { + "epoch": 9.879423328964613, + "grad_norm": 0.3664274215698242, + "learning_rate": 2.223684213745192e-08, + "loss": 0.0636, + "num_input_tokens_seen": 10104904, + "step": 37690 + }, + { + "epoch": 9.880733944954128, + "grad_norm": 3.5555660724639893, + "learning_rate": 2.1757166575425702e-08, + "loss": 0.1645, + "num_input_tokens_seen": 10106328, + "step": 37695 + }, + { + "epoch": 9.882044560943644, + "grad_norm": 1.7373945713043213, + "learning_rate": 2.1282718916465494e-08, + "loss": 0.0896, + "num_input_tokens_seen": 10107496, + "step": 37700 + }, + { + "epoch": 9.883355176933158, + "grad_norm": 1.3375728130340576, + "learning_rate": 2.0813499259872414e-08, + "loss": 0.1148, + "num_input_tokens_seen": 10108632, + "step": 37705 + }, + { + "epoch": 9.884665792922673, + "grad_norm": 2.064788579940796, + "learning_rate": 2.0349507703851244e-08, + "loss": 0.1435, + "num_input_tokens_seen": 10110008, + "step": 37710 + }, + { + "epoch": 9.885976408912189, + "grad_norm": 2.1630563735961914, + "learning_rate": 1.9890744345518742e-08, + "loss": 0.174, + "num_input_tokens_seen": 10111272, + "step": 37715 + }, + { + "epoch": 9.887287024901704, + "grad_norm": 1.6540876626968384, + "learning_rate": 1.9437209280889768e-08, + "loss": 0.1573, + "num_input_tokens_seen": 10112920, + "step": 37720 + }, + { + "epoch": 9.888597640891218, + "grad_norm": 1.3307443857192993, + "learning_rate": 1.8988902604891166e-08, + "loss": 0.3003, + "num_input_tokens_seen": 10114760, + "step": 37725 + }, + { + "epoch": 9.889908256880734, + "grad_norm": 1.5915641784667969, + "learning_rate": 1.8545824411350665e-08, + "loss": 0.2854, + "num_input_tokens_seen": 10116120, + "step": 37730 + }, + { + "epoch": 9.89121887287025, + "grad_norm": 1.5551161766052246, + "learning_rate": 1.810797479300519e-08, + "loss": 0.1543, + "num_input_tokens_seen": 10117784, + "step": 37735 + }, + { + "epoch": 9.892529488859765, + "grad_norm": 1.4364087581634521, + "learning_rate": 1.7675353841495325e-08, + "loss": 0.1859, + "num_input_tokens_seen": 10119112, + "step": 37740 + }, + { + "epoch": 9.893840104849279, + "grad_norm": 2.7466349601745605, + "learning_rate": 1.7247961647368082e-08, + "loss": 0.2088, + "num_input_tokens_seen": 10120424, + "step": 37745 + }, + { + "epoch": 9.895150720838794, + "grad_norm": 2.6359949111938477, + "learning_rate": 1.6825798300074137e-08, + "loss": 0.102, + "num_input_tokens_seen": 10121592, + "step": 37750 + }, + { + "epoch": 9.89646133682831, + "grad_norm": 0.5974667072296143, + "learning_rate": 1.640886388797336e-08, + "loss": 0.0713, + "num_input_tokens_seen": 10122600, + "step": 37755 + }, + { + "epoch": 9.897771952817825, + "grad_norm": 1.6585477590560913, + "learning_rate": 1.5997158498329277e-08, + "loss": 0.1229, + "num_input_tokens_seen": 10123880, + "step": 37760 + }, + { + "epoch": 9.899082568807339, + "grad_norm": 1.6093499660491943, + "learning_rate": 1.559068221731186e-08, + "loss": 0.0939, + "num_input_tokens_seen": 10125112, + "step": 37765 + }, + { + "epoch": 9.900393184796854, + "grad_norm": 1.9115962982177734, + "learning_rate": 1.518943512999471e-08, + "loss": 0.1205, + "num_input_tokens_seen": 10126280, + "step": 37770 + }, + { + "epoch": 9.90170380078637, + "grad_norm": 2.369171142578125, + "learning_rate": 1.4793417320357882e-08, + "loss": 0.1311, + "num_input_tokens_seen": 10128024, + "step": 37775 + }, + { + "epoch": 9.903014416775886, + "grad_norm": 0.8189470767974854, + "learning_rate": 1.4402628871285073e-08, + "loss": 0.1458, + "num_input_tokens_seen": 10129528, + "step": 37780 + }, + { + "epoch": 9.9043250327654, + "grad_norm": 2.515028953552246, + "learning_rate": 1.401706986457474e-08, + "loss": 0.1462, + "num_input_tokens_seen": 10131192, + "step": 37785 + }, + { + "epoch": 9.905635648754915, + "grad_norm": 4.594852924346924, + "learning_rate": 1.3636740380915113e-08, + "loss": 0.1805, + "num_input_tokens_seen": 10132424, + "step": 37790 + }, + { + "epoch": 9.90694626474443, + "grad_norm": 3.113548755645752, + "learning_rate": 1.326164049991474e-08, + "loss": 0.1156, + "num_input_tokens_seen": 10133816, + "step": 37795 + }, + { + "epoch": 9.908256880733944, + "grad_norm": 1.8284001350402832, + "learning_rate": 1.2891770300080263e-08, + "loss": 0.2247, + "num_input_tokens_seen": 10135176, + "step": 37800 + }, + { + "epoch": 9.90956749672346, + "grad_norm": 1.5402979850769043, + "learning_rate": 1.2527129858821985e-08, + "loss": 0.154, + "num_input_tokens_seen": 10136376, + "step": 37805 + }, + { + "epoch": 9.910878112712975, + "grad_norm": 1.4986555576324463, + "learning_rate": 1.2167719252462183e-08, + "loss": 0.2345, + "num_input_tokens_seen": 10138408, + "step": 37810 + }, + { + "epoch": 9.91218872870249, + "grad_norm": 3.0684561729431152, + "learning_rate": 1.1813538556221249e-08, + "loss": 0.1353, + "num_input_tokens_seen": 10139672, + "step": 37815 + }, + { + "epoch": 9.913499344692005, + "grad_norm": 2.508120059967041, + "learning_rate": 1.1464587844231544e-08, + "loss": 0.2638, + "num_input_tokens_seen": 10141624, + "step": 37820 + }, + { + "epoch": 9.91480996068152, + "grad_norm": 3.084420919418335, + "learning_rate": 1.1120867189526318e-08, + "loss": 0.1724, + "num_input_tokens_seen": 10142920, + "step": 37825 + }, + { + "epoch": 9.916120576671036, + "grad_norm": 1.3487844467163086, + "learning_rate": 1.0782376664045246e-08, + "loss": 0.1568, + "num_input_tokens_seen": 10144664, + "step": 37830 + }, + { + "epoch": 9.917431192660551, + "grad_norm": 2.6153488159179688, + "learning_rate": 1.044911633863721e-08, + "loss": 0.0871, + "num_input_tokens_seen": 10146008, + "step": 37835 + }, + { + "epoch": 9.918741808650065, + "grad_norm": 1.2215759754180908, + "learning_rate": 1.0121086283049198e-08, + "loss": 0.1192, + "num_input_tokens_seen": 10147608, + "step": 37840 + }, + { + "epoch": 9.92005242463958, + "grad_norm": 2.3624401092529297, + "learning_rate": 9.798286565937399e-09, + "loss": 0.2001, + "num_input_tokens_seen": 10148712, + "step": 37845 + }, + { + "epoch": 9.921363040629096, + "grad_norm": 2.6781089305877686, + "learning_rate": 9.48071725486721e-09, + "loss": 0.1447, + "num_input_tokens_seen": 10150040, + "step": 37850 + }, + { + "epoch": 9.922673656618612, + "grad_norm": 0.22607003152370453, + "learning_rate": 9.168378416299362e-09, + "loss": 0.1506, + "num_input_tokens_seen": 10152024, + "step": 37855 + }, + { + "epoch": 9.923984272608125, + "grad_norm": 7.649331569671631, + "learning_rate": 8.861270115612108e-09, + "loss": 0.2274, + "num_input_tokens_seen": 10153480, + "step": 37860 + }, + { + "epoch": 9.92529488859764, + "grad_norm": 1.380922794342041, + "learning_rate": 8.559392417079038e-09, + "loss": 0.1447, + "num_input_tokens_seen": 10154936, + "step": 37865 + }, + { + "epoch": 9.926605504587156, + "grad_norm": 1.7246029376983643, + "learning_rate": 8.262745383880166e-09, + "loss": 0.0527, + "num_input_tokens_seen": 10155960, + "step": 37870 + }, + { + "epoch": 9.927916120576672, + "grad_norm": 3.2846202850341797, + "learning_rate": 7.971329078110268e-09, + "loss": 0.2633, + "num_input_tokens_seen": 10157160, + "step": 37875 + }, + { + "epoch": 9.929226736566186, + "grad_norm": 1.5149073600769043, + "learning_rate": 7.68514356075667e-09, + "loss": 0.1345, + "num_input_tokens_seen": 10158472, + "step": 37880 + }, + { + "epoch": 9.930537352555701, + "grad_norm": 1.7887345552444458, + "learning_rate": 7.4041888917186775e-09, + "loss": 0.2196, + "num_input_tokens_seen": 10159832, + "step": 37885 + }, + { + "epoch": 9.931847968545217, + "grad_norm": 1.6411888599395752, + "learning_rate": 7.1284651297992556e-09, + "loss": 0.0851, + "num_input_tokens_seen": 10161096, + "step": 37890 + }, + { + "epoch": 9.93315858453473, + "grad_norm": 1.6430965662002563, + "learning_rate": 6.8579723327105715e-09, + "loss": 0.1187, + "num_input_tokens_seen": 10162616, + "step": 37895 + }, + { + "epoch": 9.934469200524246, + "grad_norm": 1.788221836090088, + "learning_rate": 6.5927105570601224e-09, + "loss": 0.1574, + "num_input_tokens_seen": 10163912, + "step": 37900 + }, + { + "epoch": 9.935779816513762, + "grad_norm": 2.2840497493743896, + "learning_rate": 6.3326798583729364e-09, + "loss": 0.1258, + "num_input_tokens_seen": 10165224, + "step": 37905 + }, + { + "epoch": 9.937090432503277, + "grad_norm": 2.613929271697998, + "learning_rate": 6.077880291069371e-09, + "loss": 0.1768, + "num_input_tokens_seen": 10166904, + "step": 37910 + }, + { + "epoch": 9.938401048492791, + "grad_norm": 2.510891914367676, + "learning_rate": 5.8283119084789895e-09, + "loss": 0.1451, + "num_input_tokens_seen": 10168280, + "step": 37915 + }, + { + "epoch": 9.939711664482306, + "grad_norm": 0.8844013810157776, + "learning_rate": 5.583974762835009e-09, + "loss": 0.1806, + "num_input_tokens_seen": 10169624, + "step": 37920 + }, + { + "epoch": 9.941022280471822, + "grad_norm": 3.650735378265381, + "learning_rate": 5.344868905279854e-09, + "loss": 0.1136, + "num_input_tokens_seen": 10170840, + "step": 37925 + }, + { + "epoch": 9.942332896461338, + "grad_norm": 0.9578921794891357, + "learning_rate": 5.110994385856826e-09, + "loss": 0.1074, + "num_input_tokens_seen": 10172008, + "step": 37930 + }, + { + "epoch": 9.943643512450851, + "grad_norm": 1.4214577674865723, + "learning_rate": 4.8823512535128845e-09, + "loss": 0.1091, + "num_input_tokens_seen": 10174936, + "step": 37935 + }, + { + "epoch": 9.944954128440367, + "grad_norm": 2.118107557296753, + "learning_rate": 4.658939556104191e-09, + "loss": 0.2523, + "num_input_tokens_seen": 10176232, + "step": 37940 + }, + { + "epoch": 9.946264744429882, + "grad_norm": 3.318709373474121, + "learning_rate": 4.440759340393341e-09, + "loss": 0.194, + "num_input_tokens_seen": 10177416, + "step": 37945 + }, + { + "epoch": 9.947575360419398, + "grad_norm": 2.352442979812622, + "learning_rate": 4.227810652041031e-09, + "loss": 0.1112, + "num_input_tokens_seen": 10178792, + "step": 37950 + }, + { + "epoch": 9.948885976408912, + "grad_norm": 1.0208834409713745, + "learning_rate": 4.0200935356171645e-09, + "loss": 0.0539, + "num_input_tokens_seen": 10179912, + "step": 37955 + }, + { + "epoch": 9.950196592398427, + "grad_norm": 1.925091028213501, + "learning_rate": 3.817608034600851e-09, + "loss": 0.0891, + "num_input_tokens_seen": 10181160, + "step": 37960 + }, + { + "epoch": 9.951507208387943, + "grad_norm": 2.6247241497039795, + "learning_rate": 3.620354191366526e-09, + "loss": 0.2526, + "num_input_tokens_seen": 10182376, + "step": 37965 + }, + { + "epoch": 9.952817824377458, + "grad_norm": 3.106609582901001, + "learning_rate": 3.4283320472033864e-09, + "loss": 0.2101, + "num_input_tokens_seen": 10183816, + "step": 37970 + }, + { + "epoch": 9.954128440366972, + "grad_norm": 2.2855451107025146, + "learning_rate": 3.241541642298729e-09, + "loss": 0.2558, + "num_input_tokens_seen": 10185016, + "step": 37975 + }, + { + "epoch": 9.955439056356488, + "grad_norm": 0.5666959881782532, + "learning_rate": 3.059983015749057e-09, + "loss": 0.1513, + "num_input_tokens_seen": 10186248, + "step": 37980 + }, + { + "epoch": 9.956749672346003, + "grad_norm": 1.4433857202529907, + "learning_rate": 2.88365620555453e-09, + "loss": 0.1296, + "num_input_tokens_seen": 10187944, + "step": 37985 + }, + { + "epoch": 9.958060288335517, + "grad_norm": 4.358205318450928, + "learning_rate": 2.712561248618961e-09, + "loss": 0.1516, + "num_input_tokens_seen": 10189064, + "step": 37990 + }, + { + "epoch": 9.959370904325032, + "grad_norm": 1.635524868965149, + "learning_rate": 2.546698180749818e-09, + "loss": 0.1585, + "num_input_tokens_seen": 10190616, + "step": 37995 + }, + { + "epoch": 9.960681520314548, + "grad_norm": 1.8142768144607544, + "learning_rate": 2.3860670366665505e-09, + "loss": 0.1679, + "num_input_tokens_seen": 10192200, + "step": 38000 + }, + { + "epoch": 9.961992136304064, + "grad_norm": 2.763930559158325, + "learning_rate": 2.2306678499867116e-09, + "loss": 0.1765, + "num_input_tokens_seen": 10193928, + "step": 38005 + }, + { + "epoch": 9.963302752293577, + "grad_norm": 2.436673641204834, + "learning_rate": 2.080500653234285e-09, + "loss": 0.151, + "num_input_tokens_seen": 10195240, + "step": 38010 + }, + { + "epoch": 9.964613368283093, + "grad_norm": 0.4303971827030182, + "learning_rate": 1.935565477839685e-09, + "loss": 0.164, + "num_input_tokens_seen": 10197112, + "step": 38015 + }, + { + "epoch": 9.965923984272608, + "grad_norm": 1.8167625665664673, + "learning_rate": 1.7958623541397546e-09, + "loss": 0.1545, + "num_input_tokens_seen": 10198408, + "step": 38020 + }, + { + "epoch": 9.967234600262124, + "grad_norm": 1.1324248313903809, + "learning_rate": 1.6613913113694424e-09, + "loss": 0.1132, + "num_input_tokens_seen": 10199784, + "step": 38025 + }, + { + "epoch": 9.968545216251638, + "grad_norm": 0.79793381690979, + "learning_rate": 1.532152377678453e-09, + "loss": 0.0771, + "num_input_tokens_seen": 10200824, + "step": 38030 + }, + { + "epoch": 9.969855832241153, + "grad_norm": 1.6503722667694092, + "learning_rate": 1.4081455801145948e-09, + "loss": 0.1052, + "num_input_tokens_seen": 10201912, + "step": 38035 + }, + { + "epoch": 9.971166448230669, + "grad_norm": 1.7685651779174805, + "learning_rate": 1.289370944629331e-09, + "loss": 0.0675, + "num_input_tokens_seen": 10202856, + "step": 38040 + }, + { + "epoch": 9.972477064220184, + "grad_norm": 2.3327901363372803, + "learning_rate": 1.1758284960861066e-09, + "loss": 0.1379, + "num_input_tokens_seen": 10203896, + "step": 38045 + }, + { + "epoch": 9.973787680209698, + "grad_norm": 1.5235193967819214, + "learning_rate": 1.0675182582464693e-09, + "loss": 0.1102, + "num_input_tokens_seen": 10205272, + "step": 38050 + }, + { + "epoch": 9.975098296199214, + "grad_norm": 2.359703302383423, + "learning_rate": 9.644402537811736e-10, + "loss": 0.2056, + "num_input_tokens_seen": 10206936, + "step": 38055 + }, + { + "epoch": 9.97640891218873, + "grad_norm": 5.442116737365723, + "learning_rate": 8.665945042618529e-10, + "loss": 0.0884, + "num_input_tokens_seen": 10207912, + "step": 38060 + }, + { + "epoch": 9.977719528178245, + "grad_norm": 1.7592191696166992, + "learning_rate": 7.739810301693462e-10, + "loss": 0.1877, + "num_input_tokens_seen": 10208904, + "step": 38065 + }, + { + "epoch": 9.979030144167758, + "grad_norm": 1.1590332984924316, + "learning_rate": 6.865998508881477e-10, + "loss": 0.2342, + "num_input_tokens_seen": 10209928, + "step": 38070 + }, + { + "epoch": 9.980340760157274, + "grad_norm": 1.947092890739441, + "learning_rate": 6.044509847064062e-10, + "loss": 0.0629, + "num_input_tokens_seen": 10211368, + "step": 38075 + }, + { + "epoch": 9.98165137614679, + "grad_norm": 2.388894557952881, + "learning_rate": 5.275344488187006e-10, + "loss": 0.1347, + "num_input_tokens_seen": 10212600, + "step": 38080 + }, + { + "epoch": 9.982961992136303, + "grad_norm": 1.7161929607391357, + "learning_rate": 4.5585025932048943e-10, + "loss": 0.1151, + "num_input_tokens_seen": 10213656, + "step": 38085 + }, + { + "epoch": 9.984272608125819, + "grad_norm": 2.514681100845337, + "learning_rate": 3.893984312164367e-10, + "loss": 0.2345, + "num_input_tokens_seen": 10215272, + "step": 38090 + }, + { + "epoch": 9.985583224115334, + "grad_norm": 0.7005771398544312, + "learning_rate": 3.28178978417637e-10, + "loss": 0.1064, + "num_input_tokens_seen": 10216776, + "step": 38095 + }, + { + "epoch": 9.98689384010485, + "grad_norm": 1.448050856590271, + "learning_rate": 2.7219191373328846e-10, + "loss": 0.0903, + "num_input_tokens_seen": 10217960, + "step": 38100 + }, + { + "epoch": 9.988204456094364, + "grad_norm": 0.7448958158493042, + "learning_rate": 2.2143724888179506e-10, + "loss": 0.0594, + "num_input_tokens_seen": 10219192, + "step": 38105 + }, + { + "epoch": 9.98951507208388, + "grad_norm": 2.385166883468628, + "learning_rate": 1.759149944879912e-10, + "loss": 0.2746, + "num_input_tokens_seen": 10220904, + "step": 38110 + }, + { + "epoch": 9.990825688073395, + "grad_norm": 2.619166612625122, + "learning_rate": 1.3562516008036597e-10, + "loss": 0.2931, + "num_input_tokens_seen": 10222152, + "step": 38115 + }, + { + "epoch": 9.99213630406291, + "grad_norm": 1.591558575630188, + "learning_rate": 1.0056775408828767e-10, + "loss": 0.1849, + "num_input_tokens_seen": 10223672, + "step": 38120 + }, + { + "epoch": 9.993446920052424, + "grad_norm": 3.4312751293182373, + "learning_rate": 7.07427838503305e-11, + "loss": 0.2286, + "num_input_tokens_seen": 10225736, + "step": 38125 + }, + { + "epoch": 9.99475753604194, + "grad_norm": 1.4629554748535156, + "learning_rate": 4.6150255611498996e-11, + "loss": 0.1213, + "num_input_tokens_seen": 10227320, + "step": 38130 + }, + { + "epoch": 9.996068152031455, + "grad_norm": 1.2809025049209595, + "learning_rate": 2.679017451490129e-11, + "loss": 0.1191, + "num_input_tokens_seen": 10228504, + "step": 38135 + }, + { + "epoch": 9.997378768020969, + "grad_norm": 3.078608512878418, + "learning_rate": 1.2662544615626993e-11, + "loss": 0.2585, + "num_input_tokens_seen": 10229624, + "step": 38140 + }, + { + "epoch": 9.998689384010484, + "grad_norm": 0.1137690395116806, + "learning_rate": 3.767368869644905e-12, + "loss": 0.2906, + "num_input_tokens_seen": 10231048, + "step": 38145 + }, + { + "epoch": 10.0, + "grad_norm": 3.0862784385681152, + "learning_rate": 1.0464913657859399e-13, + "loss": 0.2814, + "num_input_tokens_seen": 10232192, + "step": 38150 + }, + { + "epoch": 10.0, + "num_input_tokens_seen": 10232192, + "step": 38150, + "total_flos": 4.6075098113664614e+17, + "train_loss": 0.3660500510378715, + "train_runtime": 3850.562, + "train_samples_per_second": 19.813, + "train_steps_per_second": 9.908 + } + ], + "logging_steps": 5, + "max_steps": 38150, + "num_input_tokens_seen": 10232192, + "num_train_epochs": 10, + "save_steps": 1908, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 4.6075098113664614e+17, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}