| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 300.0, | |
| "eval_steps": 150, | |
| "global_step": 6900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 0.0003, | |
| "loss": 35.2887, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 0.00029781021897810217, | |
| "loss": 5.9569, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "learning_rate": 0.00029562043795620436, | |
| "loss": 4.9138, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "eval_cer": 1.0, | |
| "eval_loss": 4.7965407371521, | |
| "eval_runtime": 1.256, | |
| "eval_samples_per_second": 35.828, | |
| "eval_steps_per_second": 2.389, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "learning_rate": 0.00029343065693430656, | |
| "loss": 4.887, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "learning_rate": 0.00029124087591240875, | |
| "loss": 4.8447, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "learning_rate": 0.00028905109489051094, | |
| "loss": 4.7484, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "eval_cer": 1.0, | |
| "eval_loss": 4.608075141906738, | |
| "eval_runtime": 1.2451, | |
| "eval_samples_per_second": 36.142, | |
| "eval_steps_per_second": 2.409, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 15.22, | |
| "learning_rate": 0.00028686131386861314, | |
| "loss": 4.6529, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "learning_rate": 0.0002846715328467153, | |
| "loss": 4.6373, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 19.57, | |
| "learning_rate": 0.00028248175182481747, | |
| "loss": 4.5894, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 19.57, | |
| "eval_cer": 0.9851301115241635, | |
| "eval_loss": 4.469708442687988, | |
| "eval_runtime": 1.2325, | |
| "eval_samples_per_second": 36.51, | |
| "eval_steps_per_second": 2.434, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 21.74, | |
| "learning_rate": 0.00028029197080291966, | |
| "loss": 4.5045, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 23.91, | |
| "learning_rate": 0.00027810218978102186, | |
| "loss": 4.4076, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 26.09, | |
| "learning_rate": 0.00027591240875912405, | |
| "loss": 4.2024, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 26.09, | |
| "eval_cer": 0.9076827757125155, | |
| "eval_loss": 4.037315845489502, | |
| "eval_runtime": 1.2357, | |
| "eval_samples_per_second": 36.417, | |
| "eval_steps_per_second": 2.428, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 28.26, | |
| "learning_rate": 0.00027372262773722625, | |
| "loss": 3.8743, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 30.43, | |
| "learning_rate": 0.00027153284671532844, | |
| "loss": 3.3488, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 32.61, | |
| "learning_rate": 0.00026934306569343063, | |
| "loss": 2.7314, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 32.61, | |
| "eval_cer": 0.5340768277571252, | |
| "eval_loss": 2.5507473945617676, | |
| "eval_runtime": 1.2278, | |
| "eval_samples_per_second": 36.651, | |
| "eval_steps_per_second": 2.443, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 34.78, | |
| "learning_rate": 0.00026715328467153283, | |
| "loss": 2.1968, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 36.96, | |
| "learning_rate": 0.000264963503649635, | |
| "loss": 1.6522, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 39.13, | |
| "learning_rate": 0.0002627737226277372, | |
| "loss": 1.2293, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 39.13, | |
| "eval_cer": 0.4138785625774473, | |
| "eval_loss": 2.01461124420166, | |
| "eval_runtime": 1.2246, | |
| "eval_samples_per_second": 36.746, | |
| "eval_steps_per_second": 2.45, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 41.3, | |
| "learning_rate": 0.0002605839416058394, | |
| "loss": 0.9292, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 43.48, | |
| "learning_rate": 0.00025839416058394155, | |
| "loss": 0.7208, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 45.65, | |
| "learning_rate": 0.00025620437956204374, | |
| "loss": 0.5544, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 45.65, | |
| "eval_cer": 0.355638166047088, | |
| "eval_loss": 1.9821244478225708, | |
| "eval_runtime": 1.2073, | |
| "eval_samples_per_second": 37.275, | |
| "eval_steps_per_second": 2.485, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 47.83, | |
| "learning_rate": 0.00025401459854014594, | |
| "loss": 0.4757, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 0.00025182481751824813, | |
| "loss": 0.3895, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 52.17, | |
| "learning_rate": 0.0002496350364963503, | |
| "loss": 0.3224, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 52.17, | |
| "eval_cer": 0.3587360594795539, | |
| "eval_loss": 2.0189881324768066, | |
| "eval_runtime": 1.1983, | |
| "eval_samples_per_second": 37.554, | |
| "eval_steps_per_second": 2.504, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 54.35, | |
| "learning_rate": 0.0002474452554744525, | |
| "loss": 0.279, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 56.52, | |
| "learning_rate": 0.0002452554744525547, | |
| "loss": 0.2285, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 58.7, | |
| "learning_rate": 0.0002430656934306569, | |
| "loss": 0.1951, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 58.7, | |
| "eval_cer": 0.36121437422552666, | |
| "eval_loss": 2.1229116916656494, | |
| "eval_runtime": 1.2603, | |
| "eval_samples_per_second": 35.706, | |
| "eval_steps_per_second": 2.38, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 60.87, | |
| "learning_rate": 0.0002408759124087591, | |
| "loss": 0.1964, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 63.04, | |
| "learning_rate": 0.0002386861313868613, | |
| "loss": 0.1622, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 65.22, | |
| "learning_rate": 0.0002364963503649635, | |
| "loss": 0.1539, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 65.22, | |
| "eval_cer": 0.3469640644361834, | |
| "eval_loss": 2.111368179321289, | |
| "eval_runtime": 1.2194, | |
| "eval_samples_per_second": 36.903, | |
| "eval_steps_per_second": 2.46, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 67.39, | |
| "learning_rate": 0.00023430656934306568, | |
| "loss": 0.1492, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 69.57, | |
| "learning_rate": 0.00023211678832116788, | |
| "loss": 0.1404, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 71.74, | |
| "learning_rate": 0.00022992700729927004, | |
| "loss": 0.1165, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 71.74, | |
| "eval_cer": 0.33147459727385375, | |
| "eval_loss": 2.274796485900879, | |
| "eval_runtime": 1.1874, | |
| "eval_samples_per_second": 37.898, | |
| "eval_steps_per_second": 2.527, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 73.91, | |
| "learning_rate": 0.00022773722627737224, | |
| "loss": 0.1268, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 76.09, | |
| "learning_rate": 0.00022554744525547443, | |
| "loss": 0.1186, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 78.26, | |
| "learning_rate": 0.00022335766423357663, | |
| "loss": 0.1119, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 78.26, | |
| "eval_cer": 0.34882280049566294, | |
| "eval_loss": 2.2390518188476562, | |
| "eval_runtime": 1.3465, | |
| "eval_samples_per_second": 33.42, | |
| "eval_steps_per_second": 2.228, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 80.43, | |
| "learning_rate": 0.00022116788321167882, | |
| "loss": 0.0988, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 82.61, | |
| "learning_rate": 0.00021897810218978101, | |
| "loss": 0.112, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 84.78, | |
| "learning_rate": 0.0002167883211678832, | |
| "loss": 0.0989, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 84.78, | |
| "eval_cer": 0.3382899628252788, | |
| "eval_loss": 2.343754529953003, | |
| "eval_runtime": 1.2055, | |
| "eval_samples_per_second": 37.329, | |
| "eval_steps_per_second": 2.489, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 86.96, | |
| "learning_rate": 0.00021459854014598537, | |
| "loss": 0.097, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 89.13, | |
| "learning_rate": 0.00021240875912408757, | |
| "loss": 0.0854, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 91.3, | |
| "learning_rate": 0.00021021897810218976, | |
| "loss": 0.0915, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 91.3, | |
| "eval_cer": 0.3587360594795539, | |
| "eval_loss": 2.121840000152588, | |
| "eval_runtime": 1.2037, | |
| "eval_samples_per_second": 37.386, | |
| "eval_steps_per_second": 2.492, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 93.48, | |
| "learning_rate": 0.00020802919708029196, | |
| "loss": 0.078, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 95.65, | |
| "learning_rate": 0.00020583941605839415, | |
| "loss": 0.0857, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 97.83, | |
| "learning_rate": 0.00020364963503649632, | |
| "loss": 0.0721, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 97.83, | |
| "eval_cer": 0.35192069392812886, | |
| "eval_loss": 2.242812395095825, | |
| "eval_runtime": 1.1964, | |
| "eval_samples_per_second": 37.614, | |
| "eval_steps_per_second": 2.508, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 0.0002014598540145985, | |
| "loss": 0.0799, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 102.17, | |
| "learning_rate": 0.0001992700729927007, | |
| "loss": 0.0798, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 104.35, | |
| "learning_rate": 0.0001970802919708029, | |
| "loss": 0.0742, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 104.35, | |
| "eval_cer": 0.33643122676579923, | |
| "eval_loss": 2.229339838027954, | |
| "eval_runtime": 1.2156, | |
| "eval_samples_per_second": 37.019, | |
| "eval_steps_per_second": 2.468, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 106.52, | |
| "learning_rate": 0.0001948905109489051, | |
| "loss": 0.0692, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 108.7, | |
| "learning_rate": 0.0001927007299270073, | |
| "loss": 0.0664, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 110.87, | |
| "learning_rate": 0.00019051094890510948, | |
| "loss": 0.0629, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 110.87, | |
| "eval_cer": 0.33705080545229243, | |
| "eval_loss": 2.2878150939941406, | |
| "eval_runtime": 1.2044, | |
| "eval_samples_per_second": 37.364, | |
| "eval_steps_per_second": 2.491, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 113.04, | |
| "learning_rate": 0.00018832116788321167, | |
| "loss": 0.0619, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 115.22, | |
| "learning_rate": 0.00018613138686131387, | |
| "loss": 0.0582, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 117.39, | |
| "learning_rate": 0.00018394160583941606, | |
| "loss": 0.0495, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 117.39, | |
| "eval_cer": 0.34076827757125155, | |
| "eval_loss": 2.2671637535095215, | |
| "eval_runtime": 1.2039, | |
| "eval_samples_per_second": 37.379, | |
| "eval_steps_per_second": 2.492, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 119.57, | |
| "learning_rate": 0.00018175182481751826, | |
| "loss": 0.0614, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 121.74, | |
| "learning_rate": 0.00017956204379562042, | |
| "loss": 0.0565, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 123.91, | |
| "learning_rate": 0.00017737226277372262, | |
| "loss": 0.0466, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 123.91, | |
| "eval_cer": 0.35254027261462206, | |
| "eval_loss": 2.2532107830047607, | |
| "eval_runtime": 1.3563, | |
| "eval_samples_per_second": 33.179, | |
| "eval_steps_per_second": 2.212, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 126.09, | |
| "learning_rate": 0.00017518248175182478, | |
| "loss": 0.0465, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 128.26, | |
| "learning_rate": 0.00017299270072992698, | |
| "loss": 0.0496, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 130.43, | |
| "learning_rate": 0.00017080291970802917, | |
| "loss": 0.0424, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 130.43, | |
| "eval_cer": 0.32589838909541513, | |
| "eval_loss": 2.2844393253326416, | |
| "eval_runtime": 1.2006, | |
| "eval_samples_per_second": 37.48, | |
| "eval_steps_per_second": 2.499, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 132.61, | |
| "learning_rate": 0.00016861313868613137, | |
| "loss": 0.0483, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 134.78, | |
| "learning_rate": 0.00016642335766423356, | |
| "loss": 0.0488, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 136.96, | |
| "learning_rate": 0.00016423357664233575, | |
| "loss": 0.0446, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 136.96, | |
| "eval_cer": 0.3252788104089219, | |
| "eval_loss": 2.2763445377349854, | |
| "eval_runtime": 1.2043, | |
| "eval_samples_per_second": 37.368, | |
| "eval_steps_per_second": 2.491, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 139.13, | |
| "learning_rate": 0.00016204379562043795, | |
| "loss": 0.0424, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 141.3, | |
| "learning_rate": 0.00015985401459854014, | |
| "loss": 0.0429, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 143.48, | |
| "learning_rate": 0.00015766423357664234, | |
| "loss": 0.0411, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 143.48, | |
| "eval_cer": 0.3302354399008674, | |
| "eval_loss": 2.301079034805298, | |
| "eval_runtime": 1.345, | |
| "eval_samples_per_second": 33.458, | |
| "eval_steps_per_second": 2.231, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 145.65, | |
| "learning_rate": 0.00015547445255474453, | |
| "loss": 0.0392, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 147.83, | |
| "learning_rate": 0.00015328467153284672, | |
| "loss": 0.0426, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 0.00015109489051094892, | |
| "loss": 0.0419, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "eval_cer": 0.3420074349442379, | |
| "eval_loss": 2.320059299468994, | |
| "eval_runtime": 1.2411, | |
| "eval_samples_per_second": 36.259, | |
| "eval_steps_per_second": 2.417, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 152.17, | |
| "learning_rate": 0.00014890510948905108, | |
| "loss": 0.0386, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 154.35, | |
| "learning_rate": 0.00014671532846715328, | |
| "loss": 0.0402, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 156.52, | |
| "learning_rate": 0.00014452554744525547, | |
| "loss": 0.0333, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 156.52, | |
| "eval_cer": 0.34386617100371747, | |
| "eval_loss": 2.364445209503174, | |
| "eval_runtime": 1.2337, | |
| "eval_samples_per_second": 36.475, | |
| "eval_steps_per_second": 2.432, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 158.7, | |
| "learning_rate": 0.00014233576642335764, | |
| "loss": 0.0434, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 160.87, | |
| "learning_rate": 0.00014014598540145983, | |
| "loss": 0.0393, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 163.04, | |
| "learning_rate": 0.00013795620437956203, | |
| "loss": 0.0384, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 163.04, | |
| "eval_cer": 0.35315985130111527, | |
| "eval_loss": 2.3685200214385986, | |
| "eval_runtime": 1.2136, | |
| "eval_samples_per_second": 37.081, | |
| "eval_steps_per_second": 2.472, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 165.22, | |
| "learning_rate": 0.00013576642335766422, | |
| "loss": 0.0324, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 167.39, | |
| "learning_rate": 0.00013357664233576641, | |
| "loss": 0.0438, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 169.57, | |
| "learning_rate": 0.0001313868613138686, | |
| "loss": 0.0367, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 169.57, | |
| "eval_cer": 0.3469640644361834, | |
| "eval_loss": 2.397036552429199, | |
| "eval_runtime": 1.2259, | |
| "eval_samples_per_second": 36.708, | |
| "eval_steps_per_second": 2.447, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 171.74, | |
| "learning_rate": 0.00012919708029197077, | |
| "loss": 0.0336, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 173.91, | |
| "learning_rate": 0.00012700729927007297, | |
| "loss": 0.037, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 176.09, | |
| "learning_rate": 0.00012481751824817516, | |
| "loss": 0.0307, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 176.09, | |
| "eval_cer": 0.3308550185873606, | |
| "eval_loss": 2.3530125617980957, | |
| "eval_runtime": 1.2484, | |
| "eval_samples_per_second": 36.047, | |
| "eval_steps_per_second": 2.403, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 178.26, | |
| "learning_rate": 0.00012262773722627736, | |
| "loss": 0.0284, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 180.43, | |
| "learning_rate": 0.00012043795620437955, | |
| "loss": 0.0233, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 182.61, | |
| "learning_rate": 0.00011824817518248174, | |
| "loss": 0.0328, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 182.61, | |
| "eval_cer": 0.33147459727385375, | |
| "eval_loss": 2.3414556980133057, | |
| "eval_runtime": 1.2281, | |
| "eval_samples_per_second": 36.64, | |
| "eval_steps_per_second": 2.443, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 184.78, | |
| "learning_rate": 0.00011605839416058394, | |
| "loss": 0.0285, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 186.96, | |
| "learning_rate": 0.00011386861313868612, | |
| "loss": 0.0222, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 189.13, | |
| "learning_rate": 0.00011167883211678831, | |
| "loss": 0.0271, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 189.13, | |
| "eval_cer": 0.3308550185873606, | |
| "eval_loss": 2.4165024757385254, | |
| "eval_runtime": 1.1891, | |
| "eval_samples_per_second": 37.844, | |
| "eval_steps_per_second": 2.523, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 191.3, | |
| "learning_rate": 0.00010948905109489051, | |
| "loss": 0.0307, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 193.48, | |
| "learning_rate": 0.00010729927007299269, | |
| "loss": 0.023, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 195.65, | |
| "learning_rate": 0.00010510948905109488, | |
| "loss": 0.0213, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 195.65, | |
| "eval_cer": 0.3451053283767038, | |
| "eval_loss": 2.447828769683838, | |
| "eval_runtime": 1.1406, | |
| "eval_samples_per_second": 39.452, | |
| "eval_steps_per_second": 2.63, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 197.83, | |
| "learning_rate": 0.00010291970802919708, | |
| "loss": 0.021, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "learning_rate": 0.00010072992700729926, | |
| "loss": 0.0246, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 202.17, | |
| "learning_rate": 9.854014598540145e-05, | |
| "loss": 0.0193, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 202.17, | |
| "eval_cer": 0.355638166047088, | |
| "eval_loss": 2.524061918258667, | |
| "eval_runtime": 1.203, | |
| "eval_samples_per_second": 37.406, | |
| "eval_steps_per_second": 2.494, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 204.35, | |
| "learning_rate": 9.635036496350364e-05, | |
| "loss": 0.0223, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 206.52, | |
| "learning_rate": 9.416058394160584e-05, | |
| "loss": 0.0223, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 208.7, | |
| "learning_rate": 9.197080291970803e-05, | |
| "loss": 0.0204, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 208.7, | |
| "eval_cer": 0.34634448574969023, | |
| "eval_loss": 2.570009708404541, | |
| "eval_runtime": 1.2664, | |
| "eval_samples_per_second": 35.533, | |
| "eval_steps_per_second": 2.369, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 210.87, | |
| "learning_rate": 8.978102189781021e-05, | |
| "loss": 0.0202, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 213.04, | |
| "learning_rate": 8.759124087591239e-05, | |
| "loss": 0.0193, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 215.22, | |
| "learning_rate": 8.540145985401459e-05, | |
| "loss": 0.0185, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 215.22, | |
| "eval_cer": 0.31784386617100374, | |
| "eval_loss": 2.583724021911621, | |
| "eval_runtime": 1.2549, | |
| "eval_samples_per_second": 35.859, | |
| "eval_steps_per_second": 2.391, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 217.39, | |
| "learning_rate": 8.321167883211678e-05, | |
| "loss": 0.0191, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 219.57, | |
| "learning_rate": 8.102189781021897e-05, | |
| "loss": 0.0169, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 221.74, | |
| "learning_rate": 7.883211678832117e-05, | |
| "loss": 0.0161, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 221.74, | |
| "eval_cer": 0.33767038413878564, | |
| "eval_loss": 2.513859987258911, | |
| "eval_runtime": 1.2515, | |
| "eval_samples_per_second": 35.958, | |
| "eval_steps_per_second": 2.397, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 223.91, | |
| "learning_rate": 7.664233576642336e-05, | |
| "loss": 0.0183, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 226.09, | |
| "learning_rate": 7.445255474452554e-05, | |
| "loss": 0.0228, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 228.26, | |
| "learning_rate": 7.226277372262774e-05, | |
| "loss": 0.0167, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 228.26, | |
| "eval_cer": 0.3351920693928129, | |
| "eval_loss": 2.5287766456604004, | |
| "eval_runtime": 1.2044, | |
| "eval_samples_per_second": 37.363, | |
| "eval_steps_per_second": 2.491, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 230.43, | |
| "learning_rate": 7.007299270072992e-05, | |
| "loss": 0.0181, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 232.61, | |
| "learning_rate": 6.788321167883211e-05, | |
| "loss": 0.0144, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 234.78, | |
| "learning_rate": 6.56934306569343e-05, | |
| "loss": 0.0148, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 234.78, | |
| "eval_cer": 0.338909541511772, | |
| "eval_loss": 2.574066400527954, | |
| "eval_runtime": 1.2534, | |
| "eval_samples_per_second": 35.904, | |
| "eval_steps_per_second": 2.394, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 236.96, | |
| "learning_rate": 6.350364963503648e-05, | |
| "loss": 0.0143, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 239.13, | |
| "learning_rate": 6.131386861313868e-05, | |
| "loss": 0.0197, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 241.3, | |
| "learning_rate": 5.912408759124087e-05, | |
| "loss": 0.0141, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 241.3, | |
| "eval_cer": 0.338909541511772, | |
| "eval_loss": 2.5173895359039307, | |
| "eval_runtime": 1.1989, | |
| "eval_samples_per_second": 37.536, | |
| "eval_steps_per_second": 2.502, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 243.48, | |
| "learning_rate": 5.693430656934306e-05, | |
| "loss": 0.0165, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 245.65, | |
| "learning_rate": 5.4744525547445253e-05, | |
| "loss": 0.0127, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 247.83, | |
| "learning_rate": 5.255474452554744e-05, | |
| "loss": 0.0122, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 247.83, | |
| "eval_cer": 0.3351920693928129, | |
| "eval_loss": 2.5573315620422363, | |
| "eval_runtime": 1.2363, | |
| "eval_samples_per_second": 36.4, | |
| "eval_steps_per_second": 2.427, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 250.0, | |
| "learning_rate": 5.036496350364963e-05, | |
| "loss": 0.0135, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 252.17, | |
| "learning_rate": 4.817518248175182e-05, | |
| "loss": 0.0116, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 254.35, | |
| "learning_rate": 4.5985401459854016e-05, | |
| "loss": 0.0115, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 254.35, | |
| "eval_cer": 0.32961586121437425, | |
| "eval_loss": 2.579023838043213, | |
| "eval_runtime": 1.2327, | |
| "eval_samples_per_second": 36.506, | |
| "eval_steps_per_second": 2.434, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 256.52, | |
| "learning_rate": 4.3795620437956196e-05, | |
| "loss": 0.0141, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 258.7, | |
| "learning_rate": 4.160583941605839e-05, | |
| "loss": 0.0143, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 260.87, | |
| "learning_rate": 3.9416058394160584e-05, | |
| "loss": 0.0141, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 260.87, | |
| "eval_cer": 0.32032218091697645, | |
| "eval_loss": 2.577375888824463, | |
| "eval_runtime": 1.2321, | |
| "eval_samples_per_second": 36.524, | |
| "eval_steps_per_second": 2.435, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 263.04, | |
| "learning_rate": 3.722627737226277e-05, | |
| "loss": 0.0116, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 265.22, | |
| "learning_rate": 3.503649635036496e-05, | |
| "loss": 0.0101, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 267.39, | |
| "learning_rate": 3.284671532846715e-05, | |
| "loss": 0.0123, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 267.39, | |
| "eval_cer": 0.3308550185873606, | |
| "eval_loss": 2.614670753479004, | |
| "eval_runtime": 1.1319, | |
| "eval_samples_per_second": 39.755, | |
| "eval_steps_per_second": 2.65, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 269.57, | |
| "learning_rate": 3.065693430656934e-05, | |
| "loss": 0.0151, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 271.74, | |
| "learning_rate": 2.846715328467153e-05, | |
| "loss": 0.0099, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 273.91, | |
| "learning_rate": 2.627737226277372e-05, | |
| "loss": 0.0214, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 273.91, | |
| "eval_cer": 0.3302354399008674, | |
| "eval_loss": 2.620166778564453, | |
| "eval_runtime": 1.262, | |
| "eval_samples_per_second": 35.657, | |
| "eval_steps_per_second": 2.377, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 276.09, | |
| "learning_rate": 2.408759124087591e-05, | |
| "loss": 0.0085, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 278.26, | |
| "learning_rate": 2.1897810218978098e-05, | |
| "loss": 0.0119, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 280.43, | |
| "learning_rate": 1.9708029197080292e-05, | |
| "loss": 0.0107, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 280.43, | |
| "eval_cer": 0.32342007434944237, | |
| "eval_loss": 2.6263809204101562, | |
| "eval_runtime": 1.2547, | |
| "eval_samples_per_second": 35.867, | |
| "eval_steps_per_second": 2.391, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 282.61, | |
| "learning_rate": 1.751824817518248e-05, | |
| "loss": 0.0107, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 284.78, | |
| "learning_rate": 1.532846715328467e-05, | |
| "loss": 0.0105, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 286.96, | |
| "learning_rate": 1.313868613138686e-05, | |
| "loss": 0.0086, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 286.96, | |
| "eval_cer": 0.3215613382899628, | |
| "eval_loss": 2.607461452484131, | |
| "eval_runtime": 1.204, | |
| "eval_samples_per_second": 37.374, | |
| "eval_steps_per_second": 2.492, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 289.13, | |
| "learning_rate": 1.0948905109489049e-05, | |
| "loss": 0.0095, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 291.3, | |
| "learning_rate": 8.75912408759124e-06, | |
| "loss": 0.0108, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 293.48, | |
| "learning_rate": 6.56934306569343e-06, | |
| "loss": 0.0106, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 293.48, | |
| "eval_cer": 0.3246592317224288, | |
| "eval_loss": 2.595982789993286, | |
| "eval_runtime": 1.1323, | |
| "eval_samples_per_second": 39.741, | |
| "eval_steps_per_second": 2.649, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 295.65, | |
| "learning_rate": 4.37956204379562e-06, | |
| "loss": 0.0143, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 297.83, | |
| "learning_rate": 2.18978102189781e-06, | |
| "loss": 0.0105, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0085, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "eval_cer": 0.32403965303593557, | |
| "eval_loss": 2.5951595306396484, | |
| "eval_runtime": 1.2068, | |
| "eval_samples_per_second": 37.288, | |
| "eval_steps_per_second": 2.486, | |
| "step": 6900 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 6900, | |
| "num_train_epochs": 300, | |
| "save_steps": 150, | |
| "total_flos": 2.3112928880616276e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |