| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.427528675703858, | |
| "global_step": 100000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.0, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 3.5777, | |
| "max_norm": 11.326861381530762, | |
| "max_norm/layer0": 11.326861381530762, | |
| "mean_norm": 7.967035204172134, | |
| "mean_norm/layer0": 7.967035204172134, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.1, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2319, | |
| "max_norm": 11.350375175476074, | |
| "max_norm/layer0": 11.350375175476074, | |
| "mean_norm": 8.056939780712128, | |
| "mean_norm/layer0": 8.056939780712128, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5317416829745597, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.9134443998336792, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.034, | |
| "eval_samples_per_second": 96.716, | |
| "eval_steps_per_second": 1.934, | |
| "step": 1000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.21, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.8521, | |
| "max_norm": 11.451685905456543, | |
| "max_norm/layer0": 11.451685905456543, | |
| "mean_norm": 8.345297634601593, | |
| "mean_norm/layer0": 8.345297634601593, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5495499021526419, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.798967719078064, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0243, | |
| "eval_samples_per_second": 97.629, | |
| "eval_steps_per_second": 1.953, | |
| "step": 2000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.31, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003, | |
| "loss": 1.7879, | |
| "max_norm": 12.298419952392578, | |
| "max_norm/layer0": 12.298419952392578, | |
| "mean_norm": 8.877436935901642, | |
| "mean_norm/layer0": 8.877436935901642, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5556751467710371, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7738969326019287, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0304, | |
| "eval_samples_per_second": 97.052, | |
| "eval_steps_per_second": 1.941, | |
| "step": 3000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.42, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004, | |
| "loss": 1.7728, | |
| "max_norm": 13.492940902709961, | |
| "max_norm/layer0": 13.492940902709961, | |
| "mean_norm": 9.665014863014221, | |
| "mean_norm/layer0": 9.665014863014221, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5564187866927593, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7666442394256592, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0276, | |
| "eval_samples_per_second": 97.317, | |
| "eval_steps_per_second": 1.946, | |
| "step": 4000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.52, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0005, | |
| "loss": 1.7686, | |
| "max_norm": 15.771540641784668, | |
| "max_norm/layer0": 15.771540641784668, | |
| "mean_norm": 10.713956594467163, | |
| "mean_norm/layer0": 10.713956594467163, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5594520547945205, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7608648538589478, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0276, | |
| "eval_samples_per_second": 97.311, | |
| "eval_steps_per_second": 1.946, | |
| "step": 5000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.63, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004947368421052632, | |
| "loss": 1.7635, | |
| "max_norm": 18.436635971069336, | |
| "max_norm/layer0": 18.436635971069336, | |
| "mean_norm": 11.895205318927765, | |
| "mean_norm/layer0": 11.895205318927765, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5597651663405088, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7554824352264404, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0292, | |
| "eval_samples_per_second": 97.159, | |
| "eval_steps_per_second": 1.943, | |
| "step": 6000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.73, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004894736842105264, | |
| "loss": 1.7523, | |
| "max_norm": 21.048765182495117, | |
| "max_norm/layer0": 21.048765182495117, | |
| "mean_norm": 13.066829144954681, | |
| "mean_norm/layer0": 13.066829144954681, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5631506849315069, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.738256812095642, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0284, | |
| "eval_samples_per_second": 97.237, | |
| "eval_steps_per_second": 1.945, | |
| "step": 7000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.83, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004842105263157895, | |
| "loss": 1.7471, | |
| "max_norm": 23.563966751098633, | |
| "max_norm/layer0": 23.563966751098633, | |
| "mean_norm": 14.23173063993454, | |
| "mean_norm/layer0": 14.23173063993454, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5643248532289629, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7368353605270386, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0291, | |
| "eval_samples_per_second": 97.172, | |
| "eval_steps_per_second": 1.943, | |
| "step": 8000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 0.94, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00047894736842105264, | |
| "loss": 1.7404, | |
| "max_norm": 26.197235107421875, | |
| "max_norm/layer0": 26.197235107421875, | |
| "mean_norm": 15.391284584999084, | |
| "mean_norm/layer0": 15.391284584999084, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5658904109589041, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7276737689971924, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.029, | |
| "eval_samples_per_second": 97.186, | |
| "eval_steps_per_second": 1.944, | |
| "step": 9000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.04, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00047368421052631577, | |
| "loss": 1.728, | |
| "max_norm": 28.83698272705078, | |
| "max_norm/layer0": 28.83698272705078, | |
| "mean_norm": 16.548602163791656, | |
| "mean_norm/layer0": 16.548602163791656, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5647358121330724, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7289787530899048, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0306, | |
| "eval_samples_per_second": 97.031, | |
| "eval_steps_per_second": 1.941, | |
| "step": 10000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.15, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00046842105263157895, | |
| "loss": 1.7195, | |
| "max_norm": 31.372026443481445, | |
| "max_norm/layer0": 31.372026443481445, | |
| "mean_norm": 17.70964866876602, | |
| "mean_norm/layer0": 17.70964866876602, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5667318982387476, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7244290113449097, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0293, | |
| "eval_samples_per_second": 97.153, | |
| "eval_steps_per_second": 1.943, | |
| "step": 11000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.25, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00046315789473684214, | |
| "loss": 1.7198, | |
| "max_norm": 33.9889030456543, | |
| "max_norm/layer0": 33.9889030456543, | |
| "mean_norm": 18.86298167705536, | |
| "mean_norm/layer0": 18.86298167705536, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5671037181996086, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7230280637741089, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0314, | |
| "eval_samples_per_second": 96.953, | |
| "eval_steps_per_second": 1.939, | |
| "step": 12000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.36, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00045789473684210527, | |
| "loss": 1.7171, | |
| "max_norm": 36.61670684814453, | |
| "max_norm/layer0": 36.61670684814453, | |
| "mean_norm": 20.012963116168976, | |
| "mean_norm/layer0": 20.012963116168976, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5689432485322896, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7177398204803467, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0289, | |
| "eval_samples_per_second": 97.193, | |
| "eval_steps_per_second": 1.944, | |
| "step": 13000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.46, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00045263157894736845, | |
| "loss": 1.7185, | |
| "max_norm": 39.28107452392578, | |
| "max_norm/layer0": 39.28107452392578, | |
| "mean_norm": 21.156790494918823, | |
| "mean_norm/layer0": 21.156790494918823, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5688258317025441, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.715006709098816, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0296, | |
| "eval_samples_per_second": 97.126, | |
| "eval_steps_per_second": 1.943, | |
| "step": 14000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.56, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004473684210526316, | |
| "loss": 1.7149, | |
| "max_norm": 41.809288024902344, | |
| "max_norm/layer0": 41.809288024902344, | |
| "mean_norm": 22.29547154903412, | |
| "mean_norm/layer0": 22.29547154903412, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5694520547945205, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7125060558319092, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0282, | |
| "eval_samples_per_second": 97.259, | |
| "eval_steps_per_second": 1.945, | |
| "step": 15000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.67, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004421052631578947, | |
| "loss": 1.7105, | |
| "max_norm": 44.21694564819336, | |
| "max_norm/layer0": 44.21694564819336, | |
| "mean_norm": 23.42733907699585, | |
| "mean_norm/layer0": 23.42733907699585, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5695303326810176, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.709671139717102, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0286, | |
| "eval_samples_per_second": 97.217, | |
| "eval_steps_per_second": 1.944, | |
| "step": 16000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.77, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00043684210526315795, | |
| "loss": 1.7107, | |
| "max_norm": 46.647300720214844, | |
| "max_norm/layer0": 46.647300720214844, | |
| "mean_norm": 24.55408787727356, | |
| "mean_norm/layer0": 24.55408787727356, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5689236790606653, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7072749137878418, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0377, | |
| "eval_samples_per_second": 96.366, | |
| "eval_steps_per_second": 1.927, | |
| "step": 17000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.88, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004315789473684211, | |
| "loss": 1.7113, | |
| "max_norm": 49.0349235534668, | |
| "max_norm/layer0": 49.0349235534668, | |
| "mean_norm": 25.673280954360962, | |
| "mean_norm/layer0": 25.673280954360962, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5711937377690802, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7024654150009155, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.031, | |
| "eval_samples_per_second": 96.991, | |
| "eval_steps_per_second": 1.94, | |
| "step": 18000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 1.98, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004263157894736842, | |
| "loss": 1.7078, | |
| "max_norm": 51.330352783203125, | |
| "max_norm/layer0": 51.330352783203125, | |
| "mean_norm": 26.78341281414032, | |
| "mean_norm/layer0": 26.78341281414032, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5701956947162427, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7047913074493408, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0324, | |
| "eval_samples_per_second": 96.861, | |
| "eval_steps_per_second": 1.937, | |
| "step": 19000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.09, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00042105263157894734, | |
| "loss": 1.693, | |
| "max_norm": 53.77473449707031, | |
| "max_norm/layer0": 53.77473449707031, | |
| "mean_norm": 27.891030192375183, | |
| "mean_norm/layer0": 27.891030192375183, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5695694716242662, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7044708728790283, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0296, | |
| "eval_samples_per_second": 97.121, | |
| "eval_steps_per_second": 1.942, | |
| "step": 20000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.19, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004157894736842106, | |
| "loss": 1.6935, | |
| "max_norm": 56.13137435913086, | |
| "max_norm/layer0": 56.13137435913086, | |
| "mean_norm": 28.993399620056152, | |
| "mean_norm/layer0": 28.993399620056152, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5694716242661448, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.706821322441101, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.03, | |
| "eval_samples_per_second": 97.083, | |
| "eval_steps_per_second": 1.942, | |
| "step": 21000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.29, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004105263157894737, | |
| "loss": 1.6962, | |
| "max_norm": 58.38813018798828, | |
| "max_norm/layer0": 58.38813018798828, | |
| "mean_norm": 30.087660908699036, | |
| "mean_norm/layer0": 30.087660908699036, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5687475538160469, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7046499252319336, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0302, | |
| "eval_samples_per_second": 97.072, | |
| "eval_steps_per_second": 1.941, | |
| "step": 22000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.4, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00040526315789473684, | |
| "loss": 1.6954, | |
| "max_norm": 60.607887268066406, | |
| "max_norm/layer0": 60.607887268066406, | |
| "mean_norm": 31.172435641288757, | |
| "mean_norm/layer0": 31.172435641288757, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5706457925636008, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.7018758058547974, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0307, | |
| "eval_samples_per_second": 97.019, | |
| "eval_steps_per_second": 1.94, | |
| "step": 23000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.5, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0004, | |
| "loss": 1.6933, | |
| "max_norm": 62.92869186401367, | |
| "max_norm/layer0": 62.92869186401367, | |
| "mean_norm": 32.24555063247681, | |
| "mean_norm/layer0": 32.24555063247681, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5725440313111546, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.700171947479248, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0287, | |
| "eval_samples_per_second": 97.209, | |
| "eval_steps_per_second": 1.944, | |
| "step": 24000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.61, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00039473684210526315, | |
| "loss": 1.6942, | |
| "max_norm": 65.25504302978516, | |
| "max_norm/layer0": 65.25504302978516, | |
| "mean_norm": 33.31111395359039, | |
| "mean_norm/layer0": 33.31111395359039, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5717221135029354, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6983325481414795, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0305, | |
| "eval_samples_per_second": 97.039, | |
| "eval_steps_per_second": 1.941, | |
| "step": 25000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.71, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00038947368421052633, | |
| "loss": 1.6935, | |
| "max_norm": 67.45101928710938, | |
| "max_norm/layer0": 67.45101928710938, | |
| "mean_norm": 34.36543405056, | |
| "mean_norm/layer0": 34.36543405056, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.573013698630137, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6938215494155884, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.031, | |
| "eval_samples_per_second": 96.993, | |
| "eval_steps_per_second": 1.94, | |
| "step": 26000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.82, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00038421052631578946, | |
| "loss": 1.6928, | |
| "max_norm": 69.7523193359375, | |
| "max_norm/layer0": 69.7523193359375, | |
| "mean_norm": 35.40895915031433, | |
| "mean_norm/layer0": 35.40895915031433, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5718786692759296, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6977686882019043, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0294, | |
| "eval_samples_per_second": 97.146, | |
| "eval_steps_per_second": 1.943, | |
| "step": 27000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 2.92, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00037894736842105265, | |
| "loss": 1.6927, | |
| "max_norm": 71.84566497802734, | |
| "max_norm/layer0": 71.84566497802734, | |
| "mean_norm": 36.44334518909454, | |
| "mean_norm/layer0": 36.44334518909454, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.571545988258317, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6934936046600342, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0309, | |
| "eval_samples_per_second": 97.006, | |
| "eval_steps_per_second": 1.94, | |
| "step": 28000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.02, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003736842105263158, | |
| "loss": 1.6855, | |
| "max_norm": 73.88700866699219, | |
| "max_norm/layer0": 73.88700866699219, | |
| "mean_norm": 37.46485388278961, | |
| "mean_norm/layer0": 37.46485388278961, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5726027397260274, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.69780695438385, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0312, | |
| "eval_samples_per_second": 96.977, | |
| "eval_steps_per_second": 1.94, | |
| "step": 29000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.13, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00036842105263157896, | |
| "loss": 1.6773, | |
| "max_norm": 76.05863189697266, | |
| "max_norm/layer0": 76.05863189697266, | |
| "mean_norm": 38.48086929321289, | |
| "mean_norm/layer0": 38.48086929321289, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5731898238747554, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6950737237930298, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0321, | |
| "eval_samples_per_second": 96.894, | |
| "eval_steps_per_second": 1.938, | |
| "step": 30000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.23, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00036315789473684214, | |
| "loss": 1.6788, | |
| "max_norm": 78.17182922363281, | |
| "max_norm/layer0": 78.17182922363281, | |
| "mean_norm": 39.482818245887756, | |
| "mean_norm/layer0": 39.482818245887756, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5728375733855186, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6925665140151978, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0261, | |
| "eval_samples_per_second": 97.46, | |
| "eval_steps_per_second": 1.949, | |
| "step": 31000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.34, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003578947368421053, | |
| "loss": 1.6813, | |
| "max_norm": 80.1740493774414, | |
| "max_norm/layer0": 80.1740493774414, | |
| "mean_norm": 40.47369468212128, | |
| "mean_norm/layer0": 40.47369468212128, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.572641878669276, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6919567584991455, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0325, | |
| "eval_samples_per_second": 96.851, | |
| "eval_steps_per_second": 1.937, | |
| "step": 32000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.44, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003526315789473684, | |
| "loss": 1.6782, | |
| "max_norm": 82.25338745117188, | |
| "max_norm/layer0": 82.25338745117188, | |
| "mean_norm": 41.45106363296509, | |
| "mean_norm/layer0": 41.45106363296509, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5732681017612524, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6926295757293701, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0268, | |
| "eval_samples_per_second": 97.393, | |
| "eval_steps_per_second": 1.948, | |
| "step": 33000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.55, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003473684210526316, | |
| "loss": 1.6801, | |
| "max_norm": 84.2289047241211, | |
| "max_norm/layer0": 84.2289047241211, | |
| "mean_norm": 42.41705143451691, | |
| "mean_norm/layer0": 42.41705143451691, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5718590998043053, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6893627643585205, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0265, | |
| "eval_samples_per_second": 97.419, | |
| "eval_steps_per_second": 1.948, | |
| "step": 34000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.65, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00034210526315789477, | |
| "loss": 1.6796, | |
| "max_norm": 86.21290588378906, | |
| "max_norm/layer0": 86.21290588378906, | |
| "mean_norm": 43.37204849720001, | |
| "mean_norm/layer0": 43.37204849720001, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5727788649706458, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.688981294631958, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0277, | |
| "eval_samples_per_second": 97.302, | |
| "eval_steps_per_second": 1.946, | |
| "step": 35000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.75, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003368421052631579, | |
| "loss": 1.6768, | |
| "max_norm": 88.06742095947266, | |
| "max_norm/layer0": 88.06742095947266, | |
| "mean_norm": 44.314213514328, | |
| "mean_norm/layer0": 44.314213514328, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5721722113502935, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6881800889968872, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0296, | |
| "eval_samples_per_second": 97.122, | |
| "eval_steps_per_second": 1.942, | |
| "step": 36000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.86, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00033157894736842103, | |
| "loss": 1.6802, | |
| "max_norm": 89.83356475830078, | |
| "max_norm/layer0": 89.83356475830078, | |
| "mean_norm": 45.24382555484772, | |
| "mean_norm/layer0": 45.24382555484772, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.573228962818004, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6871685981750488, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0305, | |
| "eval_samples_per_second": 97.04, | |
| "eval_steps_per_second": 1.941, | |
| "step": 37000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 3.96, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003263157894736842, | |
| "loss": 1.6809, | |
| "max_norm": 91.73099517822266, | |
| "max_norm/layer0": 91.73099517822266, | |
| "mean_norm": 46.16047787666321, | |
| "mean_norm/layer0": 46.16047787666321, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5750097847358121, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.685491681098938, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0263, | |
| "eval_samples_per_second": 97.433, | |
| "eval_steps_per_second": 1.949, | |
| "step": 38000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.07, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003210526315789474, | |
| "loss": 1.6701, | |
| "max_norm": 93.56365966796875, | |
| "max_norm/layer0": 93.56365966796875, | |
| "mean_norm": 47.06712102890015, | |
| "mean_norm/layer0": 47.06712102890015, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5742074363992172, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6885604858398438, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0277, | |
| "eval_samples_per_second": 97.308, | |
| "eval_steps_per_second": 1.946, | |
| "step": 39000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.17, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00031578947368421053, | |
| "loss": 1.6646, | |
| "max_norm": 95.40019226074219, | |
| "max_norm/layer0": 95.40019226074219, | |
| "mean_norm": 47.9597909450531, | |
| "mean_norm/layer0": 47.9597909450531, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5734442270058708, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6889522075653076, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0311, | |
| "eval_samples_per_second": 96.985, | |
| "eval_steps_per_second": 1.94, | |
| "step": 40000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.28, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003105263157894737, | |
| "loss": 1.669, | |
| "max_norm": 97.21604919433594, | |
| "max_norm/layer0": 97.21604919433594, | |
| "mean_norm": 48.839876651763916, | |
| "mean_norm/layer0": 48.839876651763916, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5746771037181996, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6858941316604614, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0321, | |
| "eval_samples_per_second": 96.889, | |
| "eval_steps_per_second": 1.938, | |
| "step": 41000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.38, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00030526315789473684, | |
| "loss": 1.6713, | |
| "max_norm": 99.00255584716797, | |
| "max_norm/layer0": 99.00255584716797, | |
| "mean_norm": 49.7055082321167, | |
| "mean_norm/layer0": 49.7055082321167, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5739726027397261, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.686662197113037, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0316, | |
| "eval_samples_per_second": 96.933, | |
| "eval_steps_per_second": 1.939, | |
| "step": 42000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.48, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0003, | |
| "loss": 1.6693, | |
| "max_norm": 100.78987884521484, | |
| "max_norm/layer0": 100.78987884521484, | |
| "mean_norm": 50.558117628097534, | |
| "mean_norm/layer0": 50.558117628097534, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5749510763209393, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.682096004486084, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0324, | |
| "eval_samples_per_second": 96.862, | |
| "eval_steps_per_second": 1.937, | |
| "step": 43000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.59, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00029473684210526316, | |
| "loss": 1.6693, | |
| "max_norm": 102.48609161376953, | |
| "max_norm/layer0": 102.48609161376953, | |
| "mean_norm": 51.397204637527466, | |
| "mean_norm/layer0": 51.397204637527466, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5747162426614482, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6822019815444946, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0343, | |
| "eval_samples_per_second": 96.681, | |
| "eval_steps_per_second": 1.934, | |
| "step": 44000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.69, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00028947368421052634, | |
| "loss": 1.6692, | |
| "max_norm": 104.09925842285156, | |
| "max_norm/layer0": 104.09925842285156, | |
| "mean_norm": 52.224265336990356, | |
| "mean_norm/layer0": 52.224265336990356, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5745009784735812, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.680064082145691, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0301, | |
| "eval_samples_per_second": 97.078, | |
| "eval_steps_per_second": 1.942, | |
| "step": 45000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.8, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00028421052631578947, | |
| "loss": 1.6703, | |
| "max_norm": 105.74360656738281, | |
| "max_norm/layer0": 105.74360656738281, | |
| "mean_norm": 53.035457372665405, | |
| "mean_norm/layer0": 53.035457372665405, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5761448140900196, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6834497451782227, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.028, | |
| "eval_samples_per_second": 97.278, | |
| "eval_steps_per_second": 1.946, | |
| "step": 46000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 4.9, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0002789473684210526, | |
| "loss": 1.6677, | |
| "max_norm": 107.3390884399414, | |
| "max_norm/layer0": 107.3390884399414, | |
| "mean_norm": 53.83394503593445, | |
| "mean_norm/layer0": 53.83394503593445, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5756360078277887, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.681907296180725, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0321, | |
| "eval_samples_per_second": 96.892, | |
| "eval_steps_per_second": 1.938, | |
| "step": 47000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.01, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00027368421052631584, | |
| "loss": 1.6682, | |
| "max_norm": 108.92868041992188, | |
| "max_norm/layer0": 108.92868041992188, | |
| "mean_norm": 54.61986470222473, | |
| "mean_norm/layer0": 54.61986470222473, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5752250489236791, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6778249740600586, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0308, | |
| "eval_samples_per_second": 97.012, | |
| "eval_steps_per_second": 1.94, | |
| "step": 48000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.11, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00026842105263157897, | |
| "loss": 1.6547, | |
| "max_norm": 110.44733428955078, | |
| "max_norm/layer0": 110.44733428955078, | |
| "mean_norm": 55.39363622665405, | |
| "mean_norm/layer0": 55.39363622665405, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.575146771037182, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6824584007263184, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0301, | |
| "eval_samples_per_second": 97.074, | |
| "eval_steps_per_second": 1.941, | |
| "step": 49000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.21, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0002631578947368421, | |
| "loss": 1.6566, | |
| "max_norm": 112.06961059570312, | |
| "max_norm/layer0": 112.06961059570312, | |
| "mean_norm": 56.14954137802124, | |
| "mean_norm/layer0": 56.14954137802124, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5757729941291585, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6824774742126465, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0319, | |
| "eval_samples_per_second": 96.905, | |
| "eval_steps_per_second": 1.938, | |
| "step": 50000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.32, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0002578947368421053, | |
| "loss": 1.6605, | |
| "max_norm": 113.63825988769531, | |
| "max_norm/layer0": 113.63825988769531, | |
| "mean_norm": 56.89331555366516, | |
| "mean_norm/layer0": 56.89331555366516, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5745988258317025, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6813552379608154, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0311, | |
| "eval_samples_per_second": 96.986, | |
| "eval_steps_per_second": 1.94, | |
| "step": 51000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.42, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0002526315789473684, | |
| "loss": 1.6603, | |
| "max_norm": 115.20358276367188, | |
| "max_norm/layer0": 115.20358276367188, | |
| "mean_norm": 57.622037410736084, | |
| "mean_norm/layer0": 57.622037410736084, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5754598825831703, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6768248081207275, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0402, | |
| "eval_samples_per_second": 96.135, | |
| "eval_steps_per_second": 1.923, | |
| "step": 52000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.53, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0002473684210526316, | |
| "loss": 1.6595, | |
| "max_norm": 116.71245574951172, | |
| "max_norm/layer0": 116.71245574951172, | |
| "mean_norm": 58.335200548172, | |
| "mean_norm/layer0": 58.335200548172, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 5.53, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5753424657534246, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6757440567016602, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0326, | |
| "eval_samples_per_second": 96.84, | |
| "eval_steps_per_second": 1.937, | |
| "step": 53000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.63, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00024210526315789475, | |
| "loss": 1.6603, | |
| "max_norm": 118.23548889160156, | |
| "max_norm/layer0": 118.23548889160156, | |
| "mean_norm": 59.03403639793396, | |
| "mean_norm/layer0": 59.03403639793396, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5738160469667319, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6769322156906128, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0316, | |
| "eval_samples_per_second": 96.941, | |
| "eval_steps_per_second": 1.939, | |
| "step": 54000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.74, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00023684210526315788, | |
| "loss": 1.662, | |
| "max_norm": 119.70069122314453, | |
| "max_norm/layer0": 119.70069122314453, | |
| "mean_norm": 59.720083475112915, | |
| "mean_norm/layer0": 59.720083475112915, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5758708414872798, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6757923364639282, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0457, | |
| "eval_samples_per_second": 95.626, | |
| "eval_steps_per_second": 1.913, | |
| "step": 55000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.84, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00023157894736842107, | |
| "loss": 1.6602, | |
| "max_norm": 121.170654296875, | |
| "max_norm/layer0": 121.170654296875, | |
| "mean_norm": 60.39123606681824, | |
| "mean_norm/layer0": 60.39123606681824, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.57573385518591, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6770671606063843, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0283, | |
| "eval_samples_per_second": 97.247, | |
| "eval_steps_per_second": 1.945, | |
| "step": 56000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 5.94, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00022631578947368422, | |
| "loss": 1.6624, | |
| "max_norm": 122.60064697265625, | |
| "max_norm/layer0": 122.60064697265625, | |
| "mean_norm": 61.04953479766846, | |
| "mean_norm/layer0": 61.04953479766846, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5770254403131115, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6749203205108643, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0377, | |
| "eval_samples_per_second": 96.365, | |
| "eval_steps_per_second": 1.927, | |
| "step": 57000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.05, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00022105263157894735, | |
| "loss": 1.6527, | |
| "max_norm": 123.97573852539062, | |
| "max_norm/layer0": 123.97573852539062, | |
| "mean_norm": 61.69300150871277, | |
| "mean_norm/layer0": 61.69300150871277, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 6.05, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5757925636007828, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6790989637374878, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0297, | |
| "eval_samples_per_second": 97.113, | |
| "eval_steps_per_second": 1.942, | |
| "step": 58000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.15, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00021578947368421054, | |
| "loss": 1.6474, | |
| "max_norm": 125.31166076660156, | |
| "max_norm/layer0": 125.31166076660156, | |
| "mean_norm": 62.322699308395386, | |
| "mean_norm/layer0": 62.322699308395386, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5772602739726027, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.676284909248352, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0311, | |
| "eval_samples_per_second": 96.983, | |
| "eval_steps_per_second": 1.94, | |
| "step": 59000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.26, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00021052631578947367, | |
| "loss": 1.6494, | |
| "max_norm": 126.64249420166016, | |
| "max_norm/layer0": 126.64249420166016, | |
| "mean_norm": 62.93570160865784, | |
| "mean_norm/layer0": 62.93570160865784, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5760861056751467, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6764713525772095, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0321, | |
| "eval_samples_per_second": 96.893, | |
| "eval_steps_per_second": 1.938, | |
| "step": 60000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.36, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00020526315789473685, | |
| "loss": 1.6539, | |
| "max_norm": 127.96533966064453, | |
| "max_norm/layer0": 127.96533966064453, | |
| "mean_norm": 63.53509712219238, | |
| "mean_norm/layer0": 63.53509712219238, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 6.36, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5763600782778865, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6741266250610352, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0379, | |
| "eval_samples_per_second": 96.35, | |
| "eval_steps_per_second": 1.927, | |
| "step": 61000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.47, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0002, | |
| "loss": 1.6539, | |
| "max_norm": 129.2185516357422, | |
| "max_norm/layer0": 129.2185516357422, | |
| "mean_norm": 64.11901497840881, | |
| "mean_norm/layer0": 64.11901497840881, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5767710371819961, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6752326488494873, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0315, | |
| "eval_samples_per_second": 96.949, | |
| "eval_steps_per_second": 1.939, | |
| "step": 62000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.57, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00019473684210526317, | |
| "loss": 1.6529, | |
| "max_norm": 130.4375, | |
| "max_norm/layer0": 130.4375, | |
| "mean_norm": 64.6885313987732, | |
| "mean_norm/layer0": 64.6885313987732, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 6.57, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5774755381604697, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6737432479858398, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0332, | |
| "eval_samples_per_second": 96.787, | |
| "eval_steps_per_second": 1.936, | |
| "step": 63000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.67, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00018947368421052632, | |
| "loss": 1.6533, | |
| "max_norm": 131.647705078125, | |
| "max_norm/layer0": 131.647705078125, | |
| "mean_norm": 65.24243497848511, | |
| "mean_norm/layer0": 65.24243497848511, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5758317025440313, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6725146770477295, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0301, | |
| "eval_samples_per_second": 97.079, | |
| "eval_steps_per_second": 1.942, | |
| "step": 64000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.78, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00018421052631578948, | |
| "loss": 1.653, | |
| "max_norm": 132.81036376953125, | |
| "max_norm/layer0": 132.81036376953125, | |
| "mean_norm": 65.78296113014221, | |
| "mean_norm/layer0": 65.78296113014221, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 6.78, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5773581213307241, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6722198724746704, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0294, | |
| "eval_samples_per_second": 97.147, | |
| "eval_steps_per_second": 1.943, | |
| "step": 65000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.88, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00017894736842105264, | |
| "loss": 1.6522, | |
| "max_norm": 133.95559692382812, | |
| "max_norm/layer0": 133.95559692382812, | |
| "mean_norm": 66.3084762096405, | |
| "mean_norm/layer0": 66.3084762096405, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5761643835616438, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6726341247558594, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0342, | |
| "eval_samples_per_second": 96.689, | |
| "eval_steps_per_second": 1.934, | |
| "step": 66000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 6.99, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0001736842105263158, | |
| "loss": 1.6528, | |
| "max_norm": 135.03582763671875, | |
| "max_norm/layer0": 135.03582763671875, | |
| "mean_norm": 66.81927680969238, | |
| "mean_norm/layer0": 66.81927680969238, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5767710371819961, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.672642707824707, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0319, | |
| "eval_samples_per_second": 96.909, | |
| "eval_steps_per_second": 1.938, | |
| "step": 67000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.09, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00016842105263157895, | |
| "loss": 1.6439, | |
| "max_norm": 136.08602905273438, | |
| "max_norm/layer0": 136.08602905273438, | |
| "mean_norm": 67.31627178192139, | |
| "mean_norm/layer0": 67.31627178192139, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 7.09, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5770841487279843, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6727759838104248, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0316, | |
| "eval_samples_per_second": 96.933, | |
| "eval_steps_per_second": 1.939, | |
| "step": 68000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.19, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0001631578947368421, | |
| "loss": 1.6403, | |
| "max_norm": 137.1239013671875, | |
| "max_norm/layer0": 137.1239013671875, | |
| "mean_norm": 67.79701733589172, | |
| "mean_norm/layer0": 67.79701733589172, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5757534246575342, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.670316219329834, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0329, | |
| "eval_samples_per_second": 96.812, | |
| "eval_steps_per_second": 1.936, | |
| "step": 69000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.3, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00015789473684210527, | |
| "loss": 1.6447, | |
| "max_norm": 138.13624572753906, | |
| "max_norm/layer0": 138.13624572753906, | |
| "mean_norm": 68.2625687122345, | |
| "mean_norm/layer0": 68.2625687122345, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.57720156555773, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6696677207946777, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0301, | |
| "eval_samples_per_second": 97.082, | |
| "eval_steps_per_second": 1.942, | |
| "step": 70000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.4, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00015263157894736842, | |
| "loss": 1.6458, | |
| "max_norm": 139.11770629882812, | |
| "max_norm/layer0": 139.11770629882812, | |
| "mean_norm": 68.71309423446655, | |
| "mean_norm/layer0": 68.71309423446655, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5776908023483366, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6694140434265137, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0301, | |
| "eval_samples_per_second": 97.075, | |
| "eval_steps_per_second": 1.941, | |
| "step": 71000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.51, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00014736842105263158, | |
| "loss": 1.6447, | |
| "max_norm": 140.086669921875, | |
| "max_norm/layer0": 140.086669921875, | |
| "mean_norm": 69.14835000038147, | |
| "mean_norm/layer0": 69.14835000038147, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5771037181996086, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6716102361679077, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0344, | |
| "eval_samples_per_second": 96.67, | |
| "eval_steps_per_second": 1.933, | |
| "step": 72000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.61, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00014210526315789474, | |
| "loss": 1.6449, | |
| "max_norm": 141.01820373535156, | |
| "max_norm/layer0": 141.01820373535156, | |
| "mean_norm": 69.5690529346466, | |
| "mean_norm/layer0": 69.5690529346466, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 7.61, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.577945205479452, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6680197715759277, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.013, | |
| "eval_samples_per_second": 98.719, | |
| "eval_steps_per_second": 1.974, | |
| "step": 73000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.72, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00013684210526315792, | |
| "loss": 1.6458, | |
| "max_norm": 141.88795471191406, | |
| "max_norm/layer0": 141.88795471191406, | |
| "mean_norm": 69.97455215454102, | |
| "mean_norm/layer0": 69.97455215454102, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 7.72, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.577866927592955, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6682908535003662, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0282, | |
| "eval_samples_per_second": 97.262, | |
| "eval_steps_per_second": 1.945, | |
| "step": 74000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.82, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00013157894736842105, | |
| "loss": 1.6447, | |
| "max_norm": 142.72747802734375, | |
| "max_norm/layer0": 142.72747802734375, | |
| "mean_norm": 70.36574029922485, | |
| "mean_norm/layer0": 70.36574029922485, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 7.82, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5778277886497064, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6680580377578735, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0275, | |
| "eval_samples_per_second": 97.327, | |
| "eval_steps_per_second": 1.947, | |
| "step": 75000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 7.92, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0001263157894736842, | |
| "loss": 1.6451, | |
| "max_norm": 143.5332489013672, | |
| "max_norm/layer0": 143.5332489013672, | |
| "mean_norm": 70.7415566444397, | |
| "mean_norm/layer0": 70.7415566444397, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5780821917808219, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6676955223083496, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.038, | |
| "eval_samples_per_second": 96.341, | |
| "eval_steps_per_second": 1.927, | |
| "step": 76000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.03, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00012105263157894738, | |
| "loss": 1.6418, | |
| "max_norm": 144.31161499023438, | |
| "max_norm/layer0": 144.31161499023438, | |
| "mean_norm": 71.10220861434937, | |
| "mean_norm/layer0": 71.10220861434937, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 8.03, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5789041095890411, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.666453242301941, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.032, | |
| "eval_samples_per_second": 96.901, | |
| "eval_steps_per_second": 1.938, | |
| "step": 77000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.13, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00011578947368421053, | |
| "loss": 1.6361, | |
| "max_norm": 145.06106567382812, | |
| "max_norm/layer0": 145.06106567382812, | |
| "mean_norm": 71.44788241386414, | |
| "mean_norm/layer0": 71.44788241386414, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 8.13, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.577866927592955, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6684386730194092, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0237, | |
| "eval_samples_per_second": 97.688, | |
| "eval_steps_per_second": 1.954, | |
| "step": 78000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.24, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00011052631578947368, | |
| "loss": 1.636, | |
| "max_norm": 145.7824249267578, | |
| "max_norm/layer0": 145.7824249267578, | |
| "mean_norm": 71.77816247940063, | |
| "mean_norm/layer0": 71.77816247940063, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 8.24, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5785909980430528, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.668695092201233, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0276, | |
| "eval_samples_per_second": 97.318, | |
| "eval_steps_per_second": 1.946, | |
| "step": 79000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.34, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.00010526315789473683, | |
| "loss": 1.6357, | |
| "max_norm": 146.4754180908203, | |
| "max_norm/layer0": 146.4754180908203, | |
| "mean_norm": 72.09290337562561, | |
| "mean_norm/layer0": 72.09290337562561, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 8.34, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.579041095890411, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6670129299163818, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0297, | |
| "eval_samples_per_second": 97.117, | |
| "eval_steps_per_second": 1.942, | |
| "step": 80000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.45, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0001, | |
| "loss": 1.6379, | |
| "max_norm": 147.1273651123047, | |
| "max_norm/layer0": 147.1273651123047, | |
| "mean_norm": 72.39218544960022, | |
| "mean_norm/layer0": 72.39218544960022, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 8.45, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5787866927592955, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6658258438110352, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0297, | |
| "eval_samples_per_second": 97.119, | |
| "eval_steps_per_second": 1.942, | |
| "step": 81000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.55, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 9.473684210526316e-05, | |
| "loss": 1.6405, | |
| "max_norm": 147.75466918945312, | |
| "max_norm/layer0": 147.75466918945312, | |
| "mean_norm": 72.67654967308044, | |
| "mean_norm/layer0": 72.67654967308044, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 8.55, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5788454011741683, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6660892963409424, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0266, | |
| "eval_samples_per_second": 97.408, | |
| "eval_steps_per_second": 1.948, | |
| "step": 82000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.65, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 8.947368421052632e-05, | |
| "loss": 1.6378, | |
| "max_norm": 148.3541259765625, | |
| "max_norm/layer0": 148.3541259765625, | |
| "mean_norm": 72.94574618339539, | |
| "mean_norm/layer0": 72.94574618339539, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5789236790606653, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6649667024612427, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0319, | |
| "eval_samples_per_second": 96.911, | |
| "eval_steps_per_second": 1.938, | |
| "step": 83000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.76, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 8.421052631578948e-05, | |
| "loss": 1.6386, | |
| "max_norm": 148.9251251220703, | |
| "max_norm/layer0": 148.9251251220703, | |
| "mean_norm": 73.1996808052063, | |
| "mean_norm/layer0": 73.1996808052063, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5784344422700587, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.664962887763977, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0344, | |
| "eval_samples_per_second": 96.67, | |
| "eval_steps_per_second": 1.933, | |
| "step": 84000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.86, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 7.894736842105263e-05, | |
| "loss": 1.638, | |
| "max_norm": 149.44464111328125, | |
| "max_norm/layer0": 149.44464111328125, | |
| "mean_norm": 73.43817734718323, | |
| "mean_norm/layer0": 73.43817734718323, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 8.86, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5784931506849315, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6643970012664795, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0336, | |
| "eval_samples_per_second": 96.753, | |
| "eval_steps_per_second": 1.935, | |
| "step": 85000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 8.97, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 7.368421052631579e-05, | |
| "loss": 1.6374, | |
| "max_norm": 149.93634033203125, | |
| "max_norm/layer0": 149.93634033203125, | |
| "mean_norm": 73.66135931015015, | |
| "mean_norm/layer0": 73.66135931015015, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5777103718199609, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6635217666625977, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0303, | |
| "eval_samples_per_second": 97.056, | |
| "eval_steps_per_second": 1.941, | |
| "step": 86000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.07, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 6.842105263157896e-05, | |
| "loss": 1.6298, | |
| "max_norm": 150.40525817871094, | |
| "max_norm/layer0": 150.40525817871094, | |
| "mean_norm": 73.86943292617798, | |
| "mean_norm/layer0": 73.86943292617798, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 9.07, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5785127201565557, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6646850109100342, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0309, | |
| "eval_samples_per_second": 97.0, | |
| "eval_steps_per_second": 1.94, | |
| "step": 87000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.18, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 6.31578947368421e-05, | |
| "loss": 1.6302, | |
| "max_norm": 150.8313751220703, | |
| "max_norm/layer0": 150.8313751220703, | |
| "mean_norm": 74.06253480911255, | |
| "mean_norm/layer0": 74.06253480911255, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5787475538160469, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.664866328239441, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0347, | |
| "eval_samples_per_second": 96.643, | |
| "eval_steps_per_second": 1.933, | |
| "step": 88000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.28, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 5.789473684210527e-05, | |
| "loss": 1.6315, | |
| "max_norm": 151.22195434570312, | |
| "max_norm/layer0": 151.22195434570312, | |
| "mean_norm": 74.23983907699585, | |
| "mean_norm/layer0": 74.23983907699585, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 9.28, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5782191780821918, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6650762557983398, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0285, | |
| "eval_samples_per_second": 97.227, | |
| "eval_steps_per_second": 1.945, | |
| "step": 89000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.38, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 5.263157894736842e-05, | |
| "loss": 1.631, | |
| "max_norm": 151.57760620117188, | |
| "max_norm/layer0": 151.57760620117188, | |
| "mean_norm": 74.40166926383972, | |
| "mean_norm/layer0": 74.40166926383972, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 9.38, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.578825831702544, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6636165380477905, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0322, | |
| "eval_samples_per_second": 96.882, | |
| "eval_steps_per_second": 1.938, | |
| "step": 90000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.49, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 4.736842105263158e-05, | |
| "loss": 1.6316, | |
| "max_norm": 151.9007110595703, | |
| "max_norm/layer0": 151.9007110595703, | |
| "mean_norm": 74.54819416999817, | |
| "mean_norm/layer0": 74.54819416999817, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 9.49, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5781604696673189, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6626789569854736, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0209, | |
| "eval_samples_per_second": 97.952, | |
| "eval_steps_per_second": 1.959, | |
| "step": 91000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.59, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 4.210526315789474e-05, | |
| "loss": 1.6286, | |
| "max_norm": 152.19346618652344, | |
| "max_norm/layer0": 152.19346618652344, | |
| "mean_norm": 74.67919540405273, | |
| "mean_norm/layer0": 74.67919540405273, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 9.59, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5783365949119373, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.664610743522644, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0316, | |
| "eval_samples_per_second": 96.938, | |
| "eval_steps_per_second": 1.939, | |
| "step": 92000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.7, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 3.6842105263157895e-05, | |
| "loss": 1.6304, | |
| "max_norm": 152.45083618164062, | |
| "max_norm/layer0": 152.45083618164062, | |
| "mean_norm": 74.79478573799133, | |
| "mean_norm/layer0": 74.79478573799133, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5800587084148728, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6632179021835327, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0258, | |
| "eval_samples_per_second": 97.488, | |
| "eval_steps_per_second": 1.95, | |
| "step": 93000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.8, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 3.157894736842105e-05, | |
| "loss": 1.6298, | |
| "max_norm": 152.67724609375, | |
| "max_norm/layer0": 152.67724609375, | |
| "mean_norm": 74.8951530456543, | |
| "mean_norm/layer0": 74.8951530456543, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5799804305283758, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6622798442840576, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0295, | |
| "eval_samples_per_second": 97.137, | |
| "eval_steps_per_second": 1.943, | |
| "step": 94000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 9.91, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 2.631578947368421e-05, | |
| "loss": 1.6309, | |
| "max_norm": 152.86326599121094, | |
| "max_norm/layer0": 152.86326599121094, | |
| "mean_norm": 74.9801697731018, | |
| "mean_norm/layer0": 74.9801697731018, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5800391389432485, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6619502305984497, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0188, | |
| "eval_samples_per_second": 98.158, | |
| "eval_steps_per_second": 1.963, | |
| "step": 95000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 10.01, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 2.105263157894737e-05, | |
| "loss": 1.6302, | |
| "max_norm": 153.0155029296875, | |
| "max_norm/layer0": 153.0155029296875, | |
| "mean_norm": 75.04964685440063, | |
| "mean_norm/layer0": 75.04964685440063, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 10.01, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5801369863013699, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6601940393447876, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0341, | |
| "eval_samples_per_second": 96.703, | |
| "eval_steps_per_second": 1.934, | |
| "step": 96000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 10.11, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 1.5789473684210526e-05, | |
| "loss": 1.6242, | |
| "max_norm": 153.1354217529297, | |
| "max_norm/layer0": 153.1354217529297, | |
| "mean_norm": 75.10380005836487, | |
| "mean_norm/layer0": 75.10380005836487, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 10.11, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5785518590998043, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6609833240509033, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0284, | |
| "eval_samples_per_second": 97.237, | |
| "eval_steps_per_second": 1.945, | |
| "step": 97000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 10.22, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 1.6258, | |
| "max_norm": 153.22213745117188, | |
| "max_norm/layer0": 153.22213745117188, | |
| "mean_norm": 75.14245867729187, | |
| "mean_norm/layer0": 75.14245867729187, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 10.22, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5794716242661448, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.660490870475769, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0267, | |
| "eval_samples_per_second": 97.402, | |
| "eval_steps_per_second": 1.948, | |
| "step": 98000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 10.32, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 1.6234, | |
| "max_norm": 153.27365112304688, | |
| "max_norm/layer0": 153.27365112304688, | |
| "mean_norm": 75.16563892364502, | |
| "mean_norm/layer0": 75.16563892364502, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 10.32, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5790802348336594, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6604704856872559, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0384, | |
| "eval_samples_per_second": 96.298, | |
| "eval_steps_per_second": 1.926, | |
| "step": 99000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 10.43, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 1.6245, | |
| "max_norm": 153.29054260253906, | |
| "max_norm/layer0": 153.29054260253906, | |
| "mean_norm": 75.17323780059814, | |
| "mean_norm/layer0": 75.17323780059814, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "eval_MSE/layer0": 0.0, | |
| "eval_accuracy": 0.5791389432485323, | |
| "eval_dead_code_fraction/layer0": 1.0, | |
| "eval_input_norm/layer0": 0.0, | |
| "eval_loss": 1.6604058742523193, | |
| "eval_multicode_k": 1, | |
| "eval_output_norm/layer0": 0.0, | |
| "eval_runtime": 1.0339, | |
| "eval_samples_per_second": 96.717, | |
| "eval_steps_per_second": 1.934, | |
| "step": 100000 | |
| }, | |
| { | |
| "MSE": 0.0, | |
| "MSE/layer0": 0.0, | |
| "dead_code_fraction": 1.0, | |
| "dead_code_fraction/layer0": 1.0, | |
| "epoch": 10.43, | |
| "input_norm": 0.0, | |
| "input_norm/layer0": 0.0, | |
| "max_norm": 153.29054260253906, | |
| "max_norm/layer0": 153.29054260253906, | |
| "mean_norm": 75.17323780059814, | |
| "mean_norm/layer0": 75.17323780059814, | |
| "multicode_k": 1, | |
| "output_norm": 0.0, | |
| "output_norm/layer0": 0.0, | |
| "step": 100000, | |
| "total_flos": 3.714781621832909e+17, | |
| "train_loss": 1.6774777018260956, | |
| "train_runtime": 132212.7109, | |
| "train_samples_per_second": 72.61, | |
| "train_steps_per_second": 0.756 | |
| } | |
| ], | |
| "max_steps": 100000, | |
| "num_train_epochs": 11, | |
| "total_flos": 3.714781621832909e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |