| { | |
| "best_metric": 29.697292972396323, | |
| "best_model_checkpoint": "./linshoufanfork-whisper-small-nan-tw/checkpoint-6000", | |
| "epoch": 1.9286403085824495, | |
| "eval_steps": 500, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 48.035465240478516, | |
| "learning_rate": 2.5e-06, | |
| "loss": 6.8285, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 142.98660278320312, | |
| "learning_rate": 5e-06, | |
| "loss": 3.2493, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 27.529985427856445, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 2.2763, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 21.324289321899414, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7193, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 13.650047302246094, | |
| "learning_rate": 9.91697110594487e-06, | |
| "loss": 1.3177, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 14.939111709594727, | |
| "learning_rate": 9.83394221188974e-06, | |
| "loss": 1.2486, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 14.85805606842041, | |
| "learning_rate": 9.750913317834608e-06, | |
| "loss": 1.1991, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 15.52128791809082, | |
| "learning_rate": 9.667884423779476e-06, | |
| "loss": 1.1126, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 22.642026901245117, | |
| "learning_rate": 9.584855529724345e-06, | |
| "loss": 1.0437, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 16.712217330932617, | |
| "learning_rate": 9.501826635669213e-06, | |
| "loss": 0.9986, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 13.973222732543945, | |
| "learning_rate": 9.418797741614083e-06, | |
| "loss": 0.9188, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 13.404074668884277, | |
| "learning_rate": 9.335768847558952e-06, | |
| "loss": 0.9244, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 12.21960163116455, | |
| "learning_rate": 9.25273995350382e-06, | |
| "loss": 0.9018, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 14.460400581359863, | |
| "learning_rate": 9.169711059448689e-06, | |
| "loss": 0.8398, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 14.076154708862305, | |
| "learning_rate": 9.086682165393557e-06, | |
| "loss": 0.8828, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 13.204269409179688, | |
| "learning_rate": 9.003653271338426e-06, | |
| "loss": 0.8503, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 10.978958129882812, | |
| "learning_rate": 8.920624377283296e-06, | |
| "loss": 0.8198, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 13.963995933532715, | |
| "learning_rate": 8.837595483228164e-06, | |
| "loss": 0.8202, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 13.337563514709473, | |
| "learning_rate": 8.754566589173033e-06, | |
| "loss": 0.7536, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 12.711252212524414, | |
| "learning_rate": 8.671537695117903e-06, | |
| "loss": 0.7938, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_cer": 55.83411121482864, | |
| "eval_loss": 0.7767874002456665, | |
| "eval_runtime": 1802.202, | |
| "eval_samples_per_second": 2.461, | |
| "eval_steps_per_second": 0.308, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 13.947765350341797, | |
| "learning_rate": 8.588508801062771e-06, | |
| "loss": 0.7784, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 11.999704360961914, | |
| "learning_rate": 8.50547990700764e-06, | |
| "loss": 0.7646, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 12.103652954101562, | |
| "learning_rate": 8.422451012952508e-06, | |
| "loss": 0.7001, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 13.490057945251465, | |
| "learning_rate": 8.339422118897376e-06, | |
| "loss": 0.7941, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 13.992444038391113, | |
| "learning_rate": 8.256393224842247e-06, | |
| "loss": 0.6561, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 14.403618812561035, | |
| "learning_rate": 8.173364330787115e-06, | |
| "loss": 0.6618, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 12.38306713104248, | |
| "learning_rate": 8.090335436731984e-06, | |
| "loss": 0.7515, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 13.88232135772705, | |
| "learning_rate": 8.007306542676852e-06, | |
| "loss": 0.6823, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 14.506720542907715, | |
| "learning_rate": 7.92427764862172e-06, | |
| "loss": 0.662, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 11.101289749145508, | |
| "learning_rate": 7.841248754566589e-06, | |
| "loss": 0.6974, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 10.762197494506836, | |
| "learning_rate": 7.758219860511459e-06, | |
| "loss": 0.6643, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 14.123621940612793, | |
| "learning_rate": 7.675190966456327e-06, | |
| "loss": 0.6878, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 13.668756484985352, | |
| "learning_rate": 7.592162072401196e-06, | |
| "loss": 0.6102, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 13.39156723022461, | |
| "learning_rate": 7.509133178346065e-06, | |
| "loss": 0.6133, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 11.687459945678711, | |
| "learning_rate": 7.426104284290934e-06, | |
| "loss": 0.6139, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 14.004112243652344, | |
| "learning_rate": 7.343075390235803e-06, | |
| "loss": 0.6257, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 13.317120552062988, | |
| "learning_rate": 7.260046496180671e-06, | |
| "loss": 0.6616, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 13.344803810119629, | |
| "learning_rate": 7.17701760212554e-06, | |
| "loss": 0.6086, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 12.64527416229248, | |
| "learning_rate": 7.09398870807041e-06, | |
| "loss": 0.5734, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 9.928169250488281, | |
| "learning_rate": 7.0109598140152775e-06, | |
| "loss": 0.5845, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_cer": 41.15215362048273, | |
| "eval_loss": 0.5947259068489075, | |
| "eval_runtime": 1807.1895, | |
| "eval_samples_per_second": 2.455, | |
| "eval_steps_per_second": 0.307, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 13.225513458251953, | |
| "learning_rate": 6.927930919960146e-06, | |
| "loss": 0.6458, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 10.804333686828613, | |
| "learning_rate": 6.844902025905016e-06, | |
| "loss": 0.5106, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 10.24815559387207, | |
| "learning_rate": 6.7618731318498845e-06, | |
| "loss": 0.523, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 11.681272506713867, | |
| "learning_rate": 6.678844237794753e-06, | |
| "loss": 0.5585, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 10.01819133758545, | |
| "learning_rate": 6.595815343739622e-06, | |
| "loss": 0.5943, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 11.715396881103516, | |
| "learning_rate": 6.512786449684491e-06, | |
| "loss": 0.5472, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 10.607870101928711, | |
| "learning_rate": 6.429757555629359e-06, | |
| "loss": 0.5579, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 12.249415397644043, | |
| "learning_rate": 6.3467286615742285e-06, | |
| "loss": 0.5269, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 12.76510238647461, | |
| "learning_rate": 6.263699767519097e-06, | |
| "loss": 0.5273, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 8.935369491577148, | |
| "learning_rate": 6.180670873463966e-06, | |
| "loss": 0.525, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 11.15725040435791, | |
| "learning_rate": 6.097641979408835e-06, | |
| "loss": 0.4792, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 11.681845664978027, | |
| "learning_rate": 6.014613085353703e-06, | |
| "loss": 0.5462, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 13.019536972045898, | |
| "learning_rate": 5.931584191298572e-06, | |
| "loss": 0.4996, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 11.789406776428223, | |
| "learning_rate": 5.848555297243441e-06, | |
| "loss": 0.5088, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 13.898345947265625, | |
| "learning_rate": 5.765526403188309e-06, | |
| "loss": 0.5069, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 11.257216453552246, | |
| "learning_rate": 5.682497509133179e-06, | |
| "loss": 0.4942, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 11.38137149810791, | |
| "learning_rate": 5.599468615078048e-06, | |
| "loss": 0.4532, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 10.852495193481445, | |
| "learning_rate": 5.516439721022916e-06, | |
| "loss": 0.5231, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 14.178400039672852, | |
| "learning_rate": 5.4334108269677856e-06, | |
| "loss": 0.5041, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 11.18582534790039, | |
| "learning_rate": 5.350381932912654e-06, | |
| "loss": 0.459, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_cer": 37.618349113215096, | |
| "eval_loss": 0.5131608247756958, | |
| "eval_runtime": 1799.4249, | |
| "eval_samples_per_second": 2.465, | |
| "eval_steps_per_second": 0.308, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 9.844304084777832, | |
| "learning_rate": 5.267353038857523e-06, | |
| "loss": 0.5035, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 11.165616035461426, | |
| "learning_rate": 5.184324144802392e-06, | |
| "loss": 0.449, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 10.43535327911377, | |
| "learning_rate": 5.10129525074726e-06, | |
| "loss": 0.4471, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 9.737510681152344, | |
| "learning_rate": 5.0182663566921295e-06, | |
| "loss": 0.4779, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 10.221022605895996, | |
| "learning_rate": 4.935237462636998e-06, | |
| "loss": 0.4266, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 7.031712532043457, | |
| "learning_rate": 4.852208568581867e-06, | |
| "loss": 0.4355, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 10.13843822479248, | |
| "learning_rate": 4.769179674526736e-06, | |
| "loss": 0.4506, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 14.100777626037598, | |
| "learning_rate": 4.686150780471604e-06, | |
| "loss": 0.4484, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 11.218331336975098, | |
| "learning_rate": 4.603121886416473e-06, | |
| "loss": 0.4637, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 9.891203880310059, | |
| "learning_rate": 4.520092992361343e-06, | |
| "loss": 0.4142, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 9.585916519165039, | |
| "learning_rate": 4.437064098306211e-06, | |
| "loss": 0.4202, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 10.81905460357666, | |
| "learning_rate": 4.35403520425108e-06, | |
| "loss": 0.4459, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 13.257423400878906, | |
| "learning_rate": 4.271006310195949e-06, | |
| "loss": 0.448, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 9.057276725769043, | |
| "learning_rate": 4.187977416140817e-06, | |
| "loss": 0.4043, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 11.002601623535156, | |
| "learning_rate": 4.104948522085686e-06, | |
| "loss": 0.4011, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 15.421494483947754, | |
| "learning_rate": 4.021919628030555e-06, | |
| "loss": 0.4208, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 12.186066627502441, | |
| "learning_rate": 3.938890733975424e-06, | |
| "loss": 0.389, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 8.680899620056152, | |
| "learning_rate": 3.855861839920293e-06, | |
| "loss": 0.4189, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 10.597740173339844, | |
| "learning_rate": 3.7728329458651612e-06, | |
| "loss": 0.3654, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 10.675308227539062, | |
| "learning_rate": 3.6898040518100305e-06, | |
| "loss": 0.3512, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_cer": 35.404720629417255, | |
| "eval_loss": 0.4709227383136749, | |
| "eval_runtime": 1802.3985, | |
| "eval_samples_per_second": 2.461, | |
| "eval_steps_per_second": 0.308, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 9.923101425170898, | |
| "learning_rate": 3.6067751577548985e-06, | |
| "loss": 0.4361, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 9.01765251159668, | |
| "learning_rate": 3.523746263699768e-06, | |
| "loss": 0.3896, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 11.23643684387207, | |
| "learning_rate": 3.4407173696446367e-06, | |
| "loss": 0.3453, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 9.193674087524414, | |
| "learning_rate": 3.3576884755895056e-06, | |
| "loss": 0.3888, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 8.438018798828125, | |
| "learning_rate": 3.274659581534374e-06, | |
| "loss": 0.3798, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 9.949082374572754, | |
| "learning_rate": 3.191630687479243e-06, | |
| "loss": 0.3828, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 12.07507610321045, | |
| "learning_rate": 3.1086017934241117e-06, | |
| "loss": 0.4027, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 12.350488662719727, | |
| "learning_rate": 3.025572899368981e-06, | |
| "loss": 0.3791, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 12.681595802307129, | |
| "learning_rate": 2.9425440053138495e-06, | |
| "loss": 0.3863, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 13.789870262145996, | |
| "learning_rate": 2.8595151112587184e-06, | |
| "loss": 0.3688, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 11.885881423950195, | |
| "learning_rate": 2.7764862172035872e-06, | |
| "loss": 0.3661, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 10.707484245300293, | |
| "learning_rate": 2.6934573231484557e-06, | |
| "loss": 0.3731, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 12.371014595031738, | |
| "learning_rate": 2.6104284290933245e-06, | |
| "loss": 0.3651, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 12.415855407714844, | |
| "learning_rate": 2.5273995350381934e-06, | |
| "loss": 0.3529, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 12.046368598937988, | |
| "learning_rate": 2.4443706409830623e-06, | |
| "loss": 0.3565, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 10.2451810836792, | |
| "learning_rate": 2.361341746927931e-06, | |
| "loss": 0.3337, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 7.761926174163818, | |
| "learning_rate": 2.2783128528728e-06, | |
| "loss": 0.3636, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 9.736420631408691, | |
| "learning_rate": 2.1952839588176684e-06, | |
| "loss": 0.346, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 9.760013580322266, | |
| "learning_rate": 2.1122550647625377e-06, | |
| "loss": 0.3535, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 9.893476486206055, | |
| "learning_rate": 2.029226170707406e-06, | |
| "loss": 0.3758, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_cer": 33.57781037471663, | |
| "eval_loss": 0.43632233142852783, | |
| "eval_runtime": 1812.7817, | |
| "eval_samples_per_second": 2.447, | |
| "eval_steps_per_second": 0.306, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 11.384421348571777, | |
| "learning_rate": 1.946197276652275e-06, | |
| "loss": 0.3466, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 9.347311973571777, | |
| "learning_rate": 1.863168382597144e-06, | |
| "loss": 0.3558, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 9.740177154541016, | |
| "learning_rate": 1.7801394885420128e-06, | |
| "loss": 0.4067, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 10.038185119628906, | |
| "learning_rate": 1.6971105944868814e-06, | |
| "loss": 0.3431, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 8.289875984191895, | |
| "learning_rate": 1.6140817004317505e-06, | |
| "loss": 0.3821, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 11.42772388458252, | |
| "learning_rate": 1.5310528063766192e-06, | |
| "loss": 0.3611, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 8.776933670043945, | |
| "learning_rate": 1.4480239123214878e-06, | |
| "loss": 0.3241, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 14.44870376586914, | |
| "learning_rate": 1.364995018266357e-06, | |
| "loss": 0.3659, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 7.697235584259033, | |
| "learning_rate": 1.2819661242112256e-06, | |
| "loss": 0.317, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 9.33436107635498, | |
| "learning_rate": 1.1989372301560944e-06, | |
| "loss": 0.3243, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 7.878904819488525, | |
| "learning_rate": 1.115908336100963e-06, | |
| "loss": 0.2832, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 8.996261596679688, | |
| "learning_rate": 1.032879442045832e-06, | |
| "loss": 0.3585, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 10.357467651367188, | |
| "learning_rate": 9.498505479907008e-07, | |
| "loss": 0.3256, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 10.002203941345215, | |
| "learning_rate": 8.668216539355696e-07, | |
| "loss": 0.3459, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 10.587177276611328, | |
| "learning_rate": 7.837927598804385e-07, | |
| "loss": 0.3161, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 7.054004192352295, | |
| "learning_rate": 7.007638658253073e-07, | |
| "loss": 0.3362, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 10.065168380737305, | |
| "learning_rate": 6.177349717701761e-07, | |
| "loss": 0.3107, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 9.83284854888916, | |
| "learning_rate": 5.347060777150448e-07, | |
| "loss": 0.3301, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 11.811662673950195, | |
| "learning_rate": 4.5167718365991366e-07, | |
| "loss": 0.3792, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 11.07596206665039, | |
| "learning_rate": 3.686482896047825e-07, | |
| "loss": 0.3191, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_cer": 32.611014801973596, | |
| "eval_loss": 0.4216199815273285, | |
| "eval_runtime": 1822.6334, | |
| "eval_samples_per_second": 2.434, | |
| "eval_steps_per_second": 0.305, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 8.368192672729492, | |
| "learning_rate": 5.222149624305782e-06, | |
| "loss": 0.3044, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 13.727489471435547, | |
| "learning_rate": 5.181313296308397e-06, | |
| "loss": 0.3297, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 9.884183883666992, | |
| "learning_rate": 5.14047696831101e-06, | |
| "loss": 0.3753, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 10.073676109313965, | |
| "learning_rate": 5.0996406403136236e-06, | |
| "loss": 0.3247, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 9.34837532043457, | |
| "learning_rate": 5.0588043123162365e-06, | |
| "loss": 0.303, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 8.949431419372559, | |
| "learning_rate": 5.01796798431885e-06, | |
| "loss": 0.2329, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 7.911171913146973, | |
| "learning_rate": 4.977131656321464e-06, | |
| "loss": 0.2096, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 6.246947288513184, | |
| "learning_rate": 4.936295328324078e-06, | |
| "loss": 0.2208, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 7.6554059982299805, | |
| "learning_rate": 4.895459000326691e-06, | |
| "loss": 0.2508, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 9.491788864135742, | |
| "learning_rate": 4.8546226723293045e-06, | |
| "loss": 0.2274, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 9.15794849395752, | |
| "learning_rate": 4.813786344331918e-06, | |
| "loss": 0.2257, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 9.842211723327637, | |
| "learning_rate": 4.772950016334531e-06, | |
| "loss": 0.227, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 7.734405040740967, | |
| "learning_rate": 4.732113688337145e-06, | |
| "loss": 0.2207, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 8.951905250549316, | |
| "learning_rate": 4.691277360339759e-06, | |
| "loss": 0.2307, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 7.3573222160339355, | |
| "learning_rate": 4.650441032342372e-06, | |
| "loss": 0.2233, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 8.47739315032959, | |
| "learning_rate": 4.6096047043449855e-06, | |
| "loss": 0.2361, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 6.945776462554932, | |
| "learning_rate": 4.568768376347599e-06, | |
| "loss": 0.2353, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 8.67324161529541, | |
| "learning_rate": 4.527932048350212e-06, | |
| "loss": 0.2302, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 6.912210464477539, | |
| "learning_rate": 4.487095720352827e-06, | |
| "loss": 0.2292, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 8.434404373168945, | |
| "learning_rate": 4.44625939235544e-06, | |
| "loss": 0.2295, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "eval_cer": 32.49766635551407, | |
| "eval_loss": 0.42611706256866455, | |
| "eval_runtime": 1848.4809, | |
| "eval_samples_per_second": 2.4, | |
| "eval_steps_per_second": 0.3, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 9.02902603149414, | |
| "learning_rate": 4.405423064358053e-06, | |
| "loss": 0.2382, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 8.05671215057373, | |
| "learning_rate": 4.364586736360667e-06, | |
| "loss": 0.222, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 10.546473503112793, | |
| "learning_rate": 4.32375040836328e-06, | |
| "loss": 0.2131, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 8.884702682495117, | |
| "learning_rate": 4.282914080365894e-06, | |
| "loss": 0.1886, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 7.569803237915039, | |
| "learning_rate": 4.242077752368508e-06, | |
| "loss": 0.222, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 6.256328105926514, | |
| "learning_rate": 4.201241424371121e-06, | |
| "loss": 0.2083, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 6.724915027618408, | |
| "learning_rate": 4.1604050963737345e-06, | |
| "loss": 0.2452, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 11.219491004943848, | |
| "learning_rate": 4.119568768376348e-06, | |
| "loss": 0.2217, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 6.6789469718933105, | |
| "learning_rate": 4.078732440378961e-06, | |
| "loss": 0.1958, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 7.929986476898193, | |
| "learning_rate": 4.037896112381575e-06, | |
| "loss": 0.1863, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 8.032015800476074, | |
| "learning_rate": 3.997059784384189e-06, | |
| "loss": 0.2153, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 8.176934242248535, | |
| "learning_rate": 3.956223456386802e-06, | |
| "loss": 0.1971, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 10.322613716125488, | |
| "learning_rate": 3.9153871283894155e-06, | |
| "loss": 0.1991, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 7.837410926818848, | |
| "learning_rate": 3.874550800392029e-06, | |
| "loss": 0.2043, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 12.684860229492188, | |
| "learning_rate": 3.833714472394642e-06, | |
| "loss": 0.2008, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 7.476794242858887, | |
| "learning_rate": 3.7928781443972564e-06, | |
| "loss": 0.2199, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 4.45359992980957, | |
| "learning_rate": 3.7520418163998693e-06, | |
| "loss": 0.196, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 9.373842239379883, | |
| "learning_rate": 3.7112054884024835e-06, | |
| "loss": 0.1939, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 6.383950233459473, | |
| "learning_rate": 3.670369160405097e-06, | |
| "loss": 0.2066, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 5.862789154052734, | |
| "learning_rate": 3.6295328324077102e-06, | |
| "loss": 0.1806, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_cer": 31.99093212428324, | |
| "eval_loss": 0.4084797203540802, | |
| "eval_runtime": 1898.761, | |
| "eval_samples_per_second": 2.336, | |
| "eval_steps_per_second": 0.292, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 7.52218770980835, | |
| "learning_rate": 3.588696504410324e-06, | |
| "loss": 0.1955, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 8.115983963012695, | |
| "learning_rate": 3.5478601764129374e-06, | |
| "loss": 0.1802, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 10.009458541870117, | |
| "learning_rate": 3.5070238484155507e-06, | |
| "loss": 0.1866, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 8.48315715789795, | |
| "learning_rate": 3.4661875204181645e-06, | |
| "loss": 0.2193, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 7.425174713134766, | |
| "learning_rate": 3.425351192420778e-06, | |
| "loss": 0.2096, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 5.02262544631958, | |
| "learning_rate": 3.384514864423391e-06, | |
| "loss": 0.1892, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 7.6967010498046875, | |
| "learning_rate": 3.343678536426005e-06, | |
| "loss": 0.1869, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 7.067899703979492, | |
| "learning_rate": 3.3028422084286183e-06, | |
| "loss": 0.1842, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 9.152185440063477, | |
| "learning_rate": 3.262005880431232e-06, | |
| "loss": 0.1601, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 5.13536262512207, | |
| "learning_rate": 3.2211695524338455e-06, | |
| "loss": 0.1766, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 8.453483581542969, | |
| "learning_rate": 3.180333224436459e-06, | |
| "loss": 0.1775, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 8.888550758361816, | |
| "learning_rate": 3.1394968964390726e-06, | |
| "loss": 0.1966, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 6.241116046905518, | |
| "learning_rate": 3.098660568441686e-06, | |
| "loss": 0.1602, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 10.21055793762207, | |
| "learning_rate": 3.0578242404442993e-06, | |
| "loss": 0.188, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 6.382270812988281, | |
| "learning_rate": 3.016987912446913e-06, | |
| "loss": 0.1686, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 8.593984603881836, | |
| "learning_rate": 2.9761515844495264e-06, | |
| "loss": 0.1959, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 8.248409271240234, | |
| "learning_rate": 2.93531525645214e-06, | |
| "loss": 0.1848, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 7.425219535827637, | |
| "learning_rate": 2.8944789284547536e-06, | |
| "loss": 0.1711, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 6.594272613525391, | |
| "learning_rate": 2.853642600457367e-06, | |
| "loss": 0.1601, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 6.964175224304199, | |
| "learning_rate": 2.8128062724599807e-06, | |
| "loss": 0.16, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "eval_cer": 31.170822776370184, | |
| "eval_loss": 0.3913029432296753, | |
| "eval_runtime": 1863.6532, | |
| "eval_samples_per_second": 2.38, | |
| "eval_steps_per_second": 0.298, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 7.741447448730469, | |
| "learning_rate": 2.771969944462594e-06, | |
| "loss": 0.1552, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 6.088663101196289, | |
| "learning_rate": 2.7311336164652074e-06, | |
| "loss": 0.1759, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 9.178170204162598, | |
| "learning_rate": 2.690297288467821e-06, | |
| "loss": 0.1635, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 9.006258010864258, | |
| "learning_rate": 2.6494609604704345e-06, | |
| "loss": 0.1986, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 7.974513530731201, | |
| "learning_rate": 2.608624632473048e-06, | |
| "loss": 0.1781, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 8.142216682434082, | |
| "learning_rate": 2.5677883044756617e-06, | |
| "loss": 0.1509, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 5.452117919921875, | |
| "learning_rate": 2.526951976478275e-06, | |
| "loss": 0.1682, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 6.821118354797363, | |
| "learning_rate": 2.4861156484808888e-06, | |
| "loss": 0.1559, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 9.8412446975708, | |
| "learning_rate": 2.4452793204835026e-06, | |
| "loss": 0.1519, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 7.750609874725342, | |
| "learning_rate": 2.404442992486116e-06, | |
| "loss": 0.1394, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 8.334457397460938, | |
| "learning_rate": 2.3636066644887293e-06, | |
| "loss": 0.1605, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 5.586342811584473, | |
| "learning_rate": 2.322770336491343e-06, | |
| "loss": 0.1622, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 8.146045684814453, | |
| "learning_rate": 2.2819340084939564e-06, | |
| "loss": 0.175, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 5.796145915985107, | |
| "learning_rate": 2.24109768049657e-06, | |
| "loss": 0.1702, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 8.852866172790527, | |
| "learning_rate": 2.2002613524991835e-06, | |
| "loss": 0.1522, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 7.19501256942749, | |
| "learning_rate": 2.159425024501797e-06, | |
| "loss": 0.1465, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 7.024486064910889, | |
| "learning_rate": 2.1185886965044107e-06, | |
| "loss": 0.1539, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 7.009402751922607, | |
| "learning_rate": 2.077752368507024e-06, | |
| "loss": 0.1442, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 4.041311264038086, | |
| "learning_rate": 2.0369160405096374e-06, | |
| "loss": 0.151, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 8.767594337463379, | |
| "learning_rate": 1.996079712512251e-06, | |
| "loss": 0.1603, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "eval_cer": 30.38405120682758, | |
| "eval_loss": 0.383564829826355, | |
| "eval_runtime": 1857.7457, | |
| "eval_samples_per_second": 2.388, | |
| "eval_steps_per_second": 0.299, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 5.629392147064209, | |
| "learning_rate": 1.9552433845148645e-06, | |
| "loss": 0.1515, | |
| "step": 5025 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 5.018975734710693, | |
| "learning_rate": 1.914407056517478e-06, | |
| "loss": 0.157, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 6.182131767272949, | |
| "learning_rate": 1.8735707285200916e-06, | |
| "loss": 0.13, | |
| "step": 5075 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 9.000260353088379, | |
| "learning_rate": 1.8327344005227052e-06, | |
| "loss": 0.1726, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 6.853832244873047, | |
| "learning_rate": 1.7918980725253188e-06, | |
| "loss": 0.1451, | |
| "step": 5125 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 5.68117618560791, | |
| "learning_rate": 1.7510617445279321e-06, | |
| "loss": 0.1518, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 4.632532119750977, | |
| "learning_rate": 1.7102254165305457e-06, | |
| "loss": 0.144, | |
| "step": 5175 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 8.772269248962402, | |
| "learning_rate": 1.6693890885331592e-06, | |
| "loss": 0.1525, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 8.809287071228027, | |
| "learning_rate": 1.6285527605357728e-06, | |
| "loss": 0.132, | |
| "step": 5225 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 7.337480545043945, | |
| "learning_rate": 1.5877164325383862e-06, | |
| "loss": 0.1549, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 5.269392013549805, | |
| "learning_rate": 1.5468801045409997e-06, | |
| "loss": 0.1524, | |
| "step": 5275 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 7.877448558807373, | |
| "learning_rate": 1.5060437765436133e-06, | |
| "loss": 0.1421, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 6.454422950744629, | |
| "learning_rate": 1.4652074485462266e-06, | |
| "loss": 0.1377, | |
| "step": 5325 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 7.873298645019531, | |
| "learning_rate": 1.4243711205488402e-06, | |
| "loss": 0.1446, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 6.6517486572265625, | |
| "learning_rate": 1.383534792551454e-06, | |
| "loss": 0.1482, | |
| "step": 5375 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 9.937956809997559, | |
| "learning_rate": 1.3426984645540676e-06, | |
| "loss": 0.1306, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 4.228558540344238, | |
| "learning_rate": 1.301862136556681e-06, | |
| "loss": 0.1229, | |
| "step": 5425 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 4.710421085357666, | |
| "learning_rate": 1.2610258085592945e-06, | |
| "loss": 0.1374, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 4.934779644012451, | |
| "learning_rate": 1.220189480561908e-06, | |
| "loss": 0.1321, | |
| "step": 5475 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 9.244394302368164, | |
| "learning_rate": 1.1793531525645214e-06, | |
| "loss": 0.1343, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "eval_cer": 30.15735431390852, | |
| "eval_loss": 0.3783666491508484, | |
| "eval_runtime": 1871.1459, | |
| "eval_samples_per_second": 2.371, | |
| "eval_steps_per_second": 0.297, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 7.236656188964844, | |
| "learning_rate": 1.138516824567135e-06, | |
| "loss": 0.1295, | |
| "step": 5525 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 6.239099502563477, | |
| "learning_rate": 1.0976804965697485e-06, | |
| "loss": 0.1378, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 4.9148945808410645, | |
| "learning_rate": 1.056844168572362e-06, | |
| "loss": 0.1272, | |
| "step": 5575 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 7.572327136993408, | |
| "learning_rate": 1.0160078405749757e-06, | |
| "loss": 0.1405, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 6.76165771484375, | |
| "learning_rate": 9.751715125775892e-07, | |
| "loss": 0.1351, | |
| "step": 5625 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 10.984220504760742, | |
| "learning_rate": 9.343351845802026e-07, | |
| "loss": 0.1467, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 4.543166637420654, | |
| "learning_rate": 8.934988565828162e-07, | |
| "loss": 0.1175, | |
| "step": 5675 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 8.191649436950684, | |
| "learning_rate": 8.526625285854297e-07, | |
| "loss": 0.1388, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 7.215826988220215, | |
| "learning_rate": 8.118262005880432e-07, | |
| "loss": 0.1354, | |
| "step": 5725 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 5.940629959106445, | |
| "learning_rate": 7.709898725906567e-07, | |
| "loss": 0.1283, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 6.796767234802246, | |
| "learning_rate": 7.301535445932702e-07, | |
| "loss": 0.1274, | |
| "step": 5775 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 8.403697967529297, | |
| "learning_rate": 6.893172165958838e-07, | |
| "loss": 0.1333, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 6.127229690551758, | |
| "learning_rate": 6.484808885984972e-07, | |
| "loss": 0.1406, | |
| "step": 5825 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 7.16465950012207, | |
| "learning_rate": 6.076445606011108e-07, | |
| "loss": 0.132, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 5.777968406677246, | |
| "learning_rate": 5.668082326037243e-07, | |
| "loss": 0.1437, | |
| "step": 5875 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 6.021764755249023, | |
| "learning_rate": 5.259719046063379e-07, | |
| "loss": 0.1203, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 5.480493068695068, | |
| "learning_rate": 4.851355766089514e-07, | |
| "loss": 0.1398, | |
| "step": 5925 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 7.609493732452393, | |
| "learning_rate": 4.442992486115649e-07, | |
| "loss": 0.1274, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 5.910650730133057, | |
| "learning_rate": 4.034629206141784e-07, | |
| "loss": 0.1352, | |
| "step": 5975 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 4.371640682220459, | |
| "learning_rate": 3.626265926167919e-07, | |
| "loss": 0.1265, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "eval_cer": 29.697292972396323, | |
| "eval_loss": 0.37359777092933655, | |
| "eval_runtime": 1867.3275, | |
| "eval_samples_per_second": 2.376, | |
| "eval_steps_per_second": 0.297, | |
| "step": 6000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 6222, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "total_flos": 2.770419843072e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |