| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9801324503311258, | |
| "eval_steps": 10, | |
| "global_step": 37, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.06000334769487381, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6668, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_loss": 0.7461267113685608, | |
| "eval_runtime": 323.0571, | |
| "eval_samples_per_second": 0.696, | |
| "eval_steps_per_second": 0.176, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.06872428953647614, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7291, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.06414613127708435, | |
| "learning_rate": 6e-05, | |
| "loss": 0.6424, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.07566527277231216, | |
| "learning_rate": 8e-05, | |
| "loss": 0.7077, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.0865904837846756, | |
| "learning_rate": 0.0001, | |
| "loss": 0.6572, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.1051245629787445, | |
| "learning_rate": 0.00012, | |
| "loss": 0.7091, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.1256977915763855, | |
| "learning_rate": 0.00014, | |
| "loss": 0.6362, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.15082339942455292, | |
| "learning_rate": 0.00016, | |
| "loss": 0.6448, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.12878084182739258, | |
| "learning_rate": 0.00018, | |
| "loss": 0.5384, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.10468369722366333, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5084, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_loss": 0.45858803391456604, | |
| "eval_runtime": 327.4151, | |
| "eval_samples_per_second": 0.687, | |
| "eval_steps_per_second": 0.174, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.10410615801811218, | |
| "learning_rate": 0.00019932383577419432, | |
| "loss": 0.4551, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.08829599618911743, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 0.3825, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.10538941621780396, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.4109, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.10487955063581467, | |
| "learning_rate": 0.00018936326403234125, | |
| "loss": 0.3916, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.08832171559333801, | |
| "learning_rate": 0.00018354878114129367, | |
| "loss": 0.2869, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.07985535264015198, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.3073, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.05276218056678772, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 0.2549, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.07263408601284027, | |
| "learning_rate": 0.00015971585917027862, | |
| "loss": 0.2838, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.06012401357293129, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.2672, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.05461546778678894, | |
| "learning_rate": 0.0001396079766039157, | |
| "loss": 0.241, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_loss": 0.24859268963336945, | |
| "eval_runtime": 327.3187, | |
| "eval_samples_per_second": 0.687, | |
| "eval_steps_per_second": 0.174, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.057535674422979355, | |
| "learning_rate": 0.00012868032327110904, | |
| "loss": 0.2472, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.046681083738803864, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.2441, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.03651268407702446, | |
| "learning_rate": 0.00010581448289104758, | |
| "loss": 0.2368, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.03193233534693718, | |
| "learning_rate": 9.418551710895243e-05, | |
| "loss": 0.2838, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.04706702381372452, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.2752, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.03650536388158798, | |
| "learning_rate": 7.131967672889101e-05, | |
| "loss": 0.2326, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.030436988919973373, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 0.2333, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.03042667545378208, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.2543, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.028066903352737427, | |
| "learning_rate": 4.028414082972141e-05, | |
| "loss": 0.2221, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.029378369450569153, | |
| "learning_rate": 3.137583621312665e-05, | |
| "loss": 0.2553, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "eval_loss": 0.22415944933891296, | |
| "eval_runtime": 329.7004, | |
| "eval_samples_per_second": 0.682, | |
| "eval_steps_per_second": 0.173, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.03442079573869705, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.2678, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.026801805943250656, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 0.2198, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.02642691694200039, | |
| "learning_rate": 1.0636735967658784e-05, | |
| "loss": 0.2596, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.026598341763019562, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.2378, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.02618224173784256, | |
| "learning_rate": 2.6955129420176196e-06, | |
| "loss": 0.2497, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.027151504531502724, | |
| "learning_rate": 6.761642258056978e-07, | |
| "loss": 0.2273, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.026682227849960327, | |
| "learning_rate": 0.0, | |
| "loss": 0.2324, | |
| "step": 37 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 37, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 9.73086940767191e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |