| { | |
| "best_metric": 0.41607236862182617, | |
| "best_model_checkpoint": "output_dir/finetuned_best_mnli/checkpoint-12000", | |
| "epoch": 1.9149282920469362, | |
| "global_step": 23500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.983702737940026e-05, | |
| "loss": 0.8277, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "eval_accuracy": 0.7111563932755985, | |
| "eval_loss": 0.6876205801963806, | |
| "eval_runtime": 133.5092, | |
| "eval_samples_per_second": 73.516, | |
| "eval_steps_per_second": 9.19, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.9674054758800523e-05, | |
| "loss": 0.6598, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_accuracy": 0.7539480387162506, | |
| "eval_loss": 0.6054476499557495, | |
| "eval_runtime": 151.2877, | |
| "eval_samples_per_second": 64.876, | |
| "eval_steps_per_second": 8.11, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 1.9511082138200782e-05, | |
| "loss": 0.6012, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "eval_accuracy": 0.7814569536423841, | |
| "eval_loss": 0.5584900975227356, | |
| "eval_runtime": 160.2112, | |
| "eval_samples_per_second": 61.263, | |
| "eval_steps_per_second": 7.659, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.9348109517601044e-05, | |
| "loss": 0.5589, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_accuracy": 0.7909322465613856, | |
| "eval_loss": 0.5319538116455078, | |
| "eval_runtime": 160.4115, | |
| "eval_samples_per_second": 61.186, | |
| "eval_steps_per_second": 7.649, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.9185136897001307e-05, | |
| "loss": 0.5631, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_accuracy": 0.7811512990320937, | |
| "eval_loss": 0.5446802973747253, | |
| "eval_runtime": 190.5286, | |
| "eval_samples_per_second": 51.515, | |
| "eval_steps_per_second": 6.44, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 1.9022164276401566e-05, | |
| "loss": 0.5556, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_accuracy": 0.795109526235354, | |
| "eval_loss": 0.5091724991798401, | |
| "eval_runtime": 100.2906, | |
| "eval_samples_per_second": 97.866, | |
| "eval_steps_per_second": 12.234, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.8859191655801828e-05, | |
| "loss": 0.5311, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_accuracy": 0.8002037697401936, | |
| "eval_loss": 0.5131673216819763, | |
| "eval_runtime": 106.1752, | |
| "eval_samples_per_second": 92.442, | |
| "eval_steps_per_second": 11.556, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.8696219035202087e-05, | |
| "loss": 0.5119, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "eval_accuracy": 0.8132450331125828, | |
| "eval_loss": 0.47580984234809875, | |
| "eval_runtime": 84.6709, | |
| "eval_samples_per_second": 115.919, | |
| "eval_steps_per_second": 14.491, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 1.853324641460235e-05, | |
| "loss": 0.4988, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.8033622007131941, | |
| "eval_loss": 0.5015984177589417, | |
| "eval_runtime": 120.1676, | |
| "eval_samples_per_second": 81.678, | |
| "eval_steps_per_second": 10.211, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.837027379400261e-05, | |
| "loss": 0.509, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "eval_accuracy": 0.8169128884360672, | |
| "eval_loss": 0.4686170220375061, | |
| "eval_runtime": 115.84, | |
| "eval_samples_per_second": 84.729, | |
| "eval_steps_per_second": 10.592, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 1.820730117340287e-05, | |
| "loss": 0.5033, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "eval_accuracy": 0.8121242995415181, | |
| "eval_loss": 0.4700980484485626, | |
| "eval_runtime": 103.4737, | |
| "eval_samples_per_second": 94.855, | |
| "eval_steps_per_second": 11.858, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.804432855280313e-05, | |
| "loss": 0.4834, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "eval_accuracy": 0.8227203260315843, | |
| "eval_loss": 0.4637387990951538, | |
| "eval_runtime": 110.8909, | |
| "eval_samples_per_second": 88.51, | |
| "eval_steps_per_second": 11.065, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 1.788135593220339e-05, | |
| "loss": 0.4863, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.8237391747325522, | |
| "eval_loss": 0.45791763067245483, | |
| "eval_runtime": 95.3317, | |
| "eval_samples_per_second": 102.956, | |
| "eval_steps_per_second": 12.871, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 1.771838331160365e-05, | |
| "loss": 0.4829, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_accuracy": 0.823331635252165, | |
| "eval_loss": 0.463431179523468, | |
| "eval_runtime": 94.439, | |
| "eval_samples_per_second": 103.93, | |
| "eval_steps_per_second": 12.993, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.7555410691003914e-05, | |
| "loss": 0.4819, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "eval_accuracy": 0.8260825267447784, | |
| "eval_loss": 0.4441950023174286, | |
| "eval_runtime": 106.424, | |
| "eval_samples_per_second": 92.225, | |
| "eval_steps_per_second": 11.529, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.7392438070404173e-05, | |
| "loss": 0.4763, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "eval_accuracy": 0.8212939378502292, | |
| "eval_loss": 0.44791945815086365, | |
| "eval_runtime": 88.0696, | |
| "eval_samples_per_second": 111.446, | |
| "eval_steps_per_second": 13.932, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.7229465449804435e-05, | |
| "loss": 0.4738, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_accuracy": 0.8275089149261334, | |
| "eval_loss": 0.45088809728622437, | |
| "eval_runtime": 92.5901, | |
| "eval_samples_per_second": 106.005, | |
| "eval_steps_per_second": 13.252, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.7066492829204694e-05, | |
| "loss": 0.4687, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "eval_accuracy": 0.8311767702496179, | |
| "eval_loss": 0.43536150455474854, | |
| "eval_runtime": 126.1698, | |
| "eval_samples_per_second": 77.792, | |
| "eval_steps_per_second": 9.725, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.6903520208604957e-05, | |
| "loss": 0.468, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_accuracy": 0.8322975038206827, | |
| "eval_loss": 0.43157872557640076, | |
| "eval_runtime": 113.9939, | |
| "eval_samples_per_second": 86.101, | |
| "eval_steps_per_second": 10.764, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.6740547588005215e-05, | |
| "loss": 0.4529, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "eval_accuracy": 0.8317880794701987, | |
| "eval_loss": 0.42749524116516113, | |
| "eval_runtime": 115.1753, | |
| "eval_samples_per_second": 85.218, | |
| "eval_steps_per_second": 10.653, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 1.6577574967405478e-05, | |
| "loss": 0.4617, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "eval_accuracy": 0.8309730005094244, | |
| "eval_loss": 0.43376246094703674, | |
| "eval_runtime": 139.6786, | |
| "eval_samples_per_second": 70.268, | |
| "eval_steps_per_second": 8.784, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.6414602346805737e-05, | |
| "loss": 0.4534, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_accuracy": 0.8359653591441671, | |
| "eval_loss": 0.42552220821380615, | |
| "eval_runtime": 96.9513, | |
| "eval_samples_per_second": 101.236, | |
| "eval_steps_per_second": 12.656, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 1.6251629726206e-05, | |
| "loss": 0.4536, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_accuracy": 0.8359653591441671, | |
| "eval_loss": 0.41881263256073, | |
| "eval_runtime": 108.0698, | |
| "eval_samples_per_second": 90.821, | |
| "eval_steps_per_second": 11.354, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 1.608865710560626e-05, | |
| "loss": 0.4551, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "eval_accuracy": 0.8391237901171676, | |
| "eval_loss": 0.41607236862182617, | |
| "eval_runtime": 89.8988, | |
| "eval_samples_per_second": 109.178, | |
| "eval_steps_per_second": 13.649, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 1.592568448500652e-05, | |
| "loss": 0.4006, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.8356597045338767, | |
| "eval_loss": 0.4454086124897003, | |
| "eval_runtime": 91.709, | |
| "eval_samples_per_second": 107.023, | |
| "eval_steps_per_second": 13.379, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 1.576271186440678e-05, | |
| "loss": 0.3355, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "eval_accuracy": 0.840855832908813, | |
| "eval_loss": 0.44230031967163086, | |
| "eval_runtime": 93.3849, | |
| "eval_samples_per_second": 105.103, | |
| "eval_steps_per_second": 13.139, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 1.5599739243807042e-05, | |
| "loss": 0.3309, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_accuracy": 0.8339276617422313, | |
| "eval_loss": 0.4691074788570404, | |
| "eval_runtime": 81.8679, | |
| "eval_samples_per_second": 119.888, | |
| "eval_steps_per_second": 14.988, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 1.54367666232073e-05, | |
| "loss": 0.3416, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_accuracy": 0.8397350993377484, | |
| "eval_loss": 0.4406871497631073, | |
| "eval_runtime": 80.3421, | |
| "eval_samples_per_second": 122.165, | |
| "eval_steps_per_second": 15.272, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 1.5273794002607563e-05, | |
| "loss": 0.3463, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "eval_accuracy": 0.8371879775853286, | |
| "eval_loss": 0.445261150598526, | |
| "eval_runtime": 83.1773, | |
| "eval_samples_per_second": 118.001, | |
| "eval_steps_per_second": 14.752, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.5110821382007822e-05, | |
| "loss": 0.3447, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "eval_accuracy": 0.8397350993377484, | |
| "eval_loss": 0.4480521082878113, | |
| "eval_runtime": 93.0739, | |
| "eval_samples_per_second": 105.454, | |
| "eval_steps_per_second": 13.183, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 1.4947848761408083e-05, | |
| "loss": 0.3302, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_accuracy": 0.8403464085583291, | |
| "eval_loss": 0.462993860244751, | |
| "eval_runtime": 98.4254, | |
| "eval_samples_per_second": 99.72, | |
| "eval_steps_per_second": 12.466, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.4784876140808346e-05, | |
| "loss": 0.3366, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "eval_accuracy": 0.8402445236882323, | |
| "eval_loss": 0.4278508424758911, | |
| "eval_runtime": 80.9392, | |
| "eval_samples_per_second": 121.264, | |
| "eval_steps_per_second": 15.16, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 1.4621903520208606e-05, | |
| "loss": 0.3446, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "eval_accuracy": 0.8444218033622007, | |
| "eval_loss": 0.4328218996524811, | |
| "eval_runtime": 88.6359, | |
| "eval_samples_per_second": 110.734, | |
| "eval_steps_per_second": 13.843, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 1.4458930899608867e-05, | |
| "loss": 0.3442, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_accuracy": 0.8421803362200713, | |
| "eval_loss": 0.4372538924217224, | |
| "eval_runtime": 90.5486, | |
| "eval_samples_per_second": 108.395, | |
| "eval_steps_per_second": 13.551, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 1.4295958279009128e-05, | |
| "loss": 0.3431, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "eval_accuracy": 0.8425878757004585, | |
| "eval_loss": 0.444231778383255, | |
| "eval_runtime": 104.7466, | |
| "eval_samples_per_second": 93.702, | |
| "eval_steps_per_second": 11.714, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 1.4132985658409388e-05, | |
| "loss": 0.3388, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "eval_accuracy": 0.8358634742740703, | |
| "eval_loss": 0.4574286937713623, | |
| "eval_runtime": 79.912, | |
| "eval_samples_per_second": 122.823, | |
| "eval_steps_per_second": 15.354, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 1.3970013037809649e-05, | |
| "loss": 0.3403, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "eval_accuracy": 0.8423841059602649, | |
| "eval_loss": 0.4366081953048706, | |
| "eval_runtime": 80.6484, | |
| "eval_samples_per_second": 121.701, | |
| "eval_steps_per_second": 15.214, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 1.380704041720991e-05, | |
| "loss": 0.3421, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "eval_accuracy": 0.8441161487519103, | |
| "eval_loss": 0.4383392930030823, | |
| "eval_runtime": 105.6168, | |
| "eval_samples_per_second": 92.93, | |
| "eval_steps_per_second": 11.617, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 1.364406779661017e-05, | |
| "loss": 0.3405, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "eval_accuracy": 0.8411614875191035, | |
| "eval_loss": 0.4455229938030243, | |
| "eval_runtime": 99.6387, | |
| "eval_samples_per_second": 98.506, | |
| "eval_steps_per_second": 12.314, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 1.3481095176010431e-05, | |
| "loss": 0.3448, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "eval_accuracy": 0.8385124808965868, | |
| "eval_loss": 0.4386064112186432, | |
| "eval_runtime": 86.9871, | |
| "eval_samples_per_second": 112.833, | |
| "eval_steps_per_second": 14.106, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 1.3318122555410693e-05, | |
| "loss": 0.3346, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "eval_accuracy": 0.8416709118695873, | |
| "eval_loss": 0.4354933500289917, | |
| "eval_runtime": 90.9919, | |
| "eval_samples_per_second": 107.867, | |
| "eval_steps_per_second": 13.485, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 1.3155149934810954e-05, | |
| "loss": 0.3475, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_accuracy": 0.8372898624554254, | |
| "eval_loss": 0.4339876174926758, | |
| "eval_runtime": 89.1677, | |
| "eval_samples_per_second": 110.073, | |
| "eval_steps_per_second": 13.761, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 1.2992177314211213e-05, | |
| "loss": 0.3378, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "eval_accuracy": 0.8469689251146205, | |
| "eval_loss": 0.4287373125553131, | |
| "eval_runtime": 93.5816, | |
| "eval_samples_per_second": 104.882, | |
| "eval_steps_per_second": 13.112, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 1.2829204693611474e-05, | |
| "loss": 0.3454, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_accuracy": 0.8434029546612328, | |
| "eval_loss": 0.43279626965522766, | |
| "eval_runtime": 108.513, | |
| "eval_samples_per_second": 90.45, | |
| "eval_steps_per_second": 11.307, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.2666232073011735e-05, | |
| "loss": 0.3423, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_accuracy": 0.8433010697911361, | |
| "eval_loss": 0.4425649344921112, | |
| "eval_runtime": 91.9329, | |
| "eval_samples_per_second": 106.763, | |
| "eval_steps_per_second": 13.347, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 1.2503259452411995e-05, | |
| "loss": 0.3292, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_accuracy": 0.845236882322975, | |
| "eval_loss": 0.424352765083313, | |
| "eval_runtime": 101.4887, | |
| "eval_samples_per_second": 96.71, | |
| "eval_steps_per_second": 12.09, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 1.2340286831812256e-05, | |
| "loss": 0.3363, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "eval_accuracy": 0.8494141619969434, | |
| "eval_loss": 0.4183717668056488, | |
| "eval_runtime": 104.207, | |
| "eval_samples_per_second": 94.188, | |
| "eval_steps_per_second": 11.775, | |
| "step": 23500 | |
| } | |
| ], | |
| "max_steps": 61360, | |
| "num_train_epochs": 5, | |
| "total_flos": 4.946519097615514e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |