| { | |
| "best_metric": 3.4012534618377686, | |
| "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained_last/de_mlm/de_childes_30/checkpoint-44000", | |
| "epoch": 33.04543747653023, | |
| "eval_steps": 2000, | |
| "global_step": 44000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.5020653398422832, | |
| "eval_loss": 7.472109317779541, | |
| "eval_runtime": 2.5707, | |
| "eval_samples_per_second": 1291.498, | |
| "eval_steps_per_second": 80.913, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0041306796845664, | |
| "grad_norm": 0.9529591798782349, | |
| "learning_rate": 1e-05, | |
| "loss": 7.3851, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.0041306796845664, | |
| "eval_loss": 6.4175896644592285, | |
| "eval_runtime": 2.5761, | |
| "eval_samples_per_second": 1288.76, | |
| "eval_steps_per_second": 80.742, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.50619601952685, | |
| "eval_loss": 6.3026041984558105, | |
| "eval_runtime": 2.5876, | |
| "eval_samples_per_second": 1283.023, | |
| "eval_steps_per_second": 80.382, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.008261359369133, | |
| "grad_norm": 1.5569523572921753, | |
| "learning_rate": 2e-05, | |
| "loss": 6.0706, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 6.008261359369133, | |
| "eval_loss": 6.200850486755371, | |
| "eval_runtime": 2.6505, | |
| "eval_samples_per_second": 1252.599, | |
| "eval_steps_per_second": 78.476, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 7.510326699211416, | |
| "eval_loss": 6.105844497680664, | |
| "eval_runtime": 2.5888, | |
| "eval_samples_per_second": 1282.425, | |
| "eval_steps_per_second": 80.345, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 9.0123920390537, | |
| "grad_norm": 2.3049111366271973, | |
| "learning_rate": 3e-05, | |
| "loss": 5.8733, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 9.0123920390537, | |
| "eval_loss": 6.0330352783203125, | |
| "eval_runtime": 2.5874, | |
| "eval_samples_per_second": 1283.132, | |
| "eval_steps_per_second": 80.389, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 10.514457378895981, | |
| "eval_loss": 5.968105792999268, | |
| "eval_runtime": 2.6066, | |
| "eval_samples_per_second": 1273.702, | |
| "eval_steps_per_second": 79.798, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 12.016522718738265, | |
| "grad_norm": 2.316237688064575, | |
| "learning_rate": 4e-05, | |
| "loss": 5.723, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 12.016522718738265, | |
| "eval_loss": 5.878941535949707, | |
| "eval_runtime": 2.5723, | |
| "eval_samples_per_second": 1290.696, | |
| "eval_steps_per_second": 80.863, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 13.518588058580548, | |
| "eval_loss": 5.8198137283325195, | |
| "eval_runtime": 2.5819, | |
| "eval_samples_per_second": 1285.857, | |
| "eval_steps_per_second": 80.56, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 15.020653398422832, | |
| "grad_norm": 2.299781322479248, | |
| "learning_rate": 5e-05, | |
| "loss": 5.6127, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 15.020653398422832, | |
| "eval_loss": 5.802139759063721, | |
| "eval_runtime": 2.6545, | |
| "eval_samples_per_second": 1250.713, | |
| "eval_steps_per_second": 78.358, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 16.522718738265116, | |
| "eval_loss": 5.7662129402160645, | |
| "eval_runtime": 2.5877, | |
| "eval_samples_per_second": 1282.983, | |
| "eval_steps_per_second": 80.38, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 18.0247840781074, | |
| "grad_norm": 2.8674566745758057, | |
| "learning_rate": 6e-05, | |
| "loss": 5.5325, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 18.0247840781074, | |
| "eval_loss": 5.731865406036377, | |
| "eval_runtime": 2.5813, | |
| "eval_samples_per_second": 1286.16, | |
| "eval_steps_per_second": 80.579, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 19.52684941794968, | |
| "eval_loss": 5.703192234039307, | |
| "eval_runtime": 2.5846, | |
| "eval_samples_per_second": 1284.544, | |
| "eval_steps_per_second": 80.477, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 21.028914757791963, | |
| "grad_norm": 4.573985576629639, | |
| "learning_rate": 7e-05, | |
| "loss": 5.4632, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 21.028914757791963, | |
| "eval_loss": 5.70460319519043, | |
| "eval_runtime": 2.5824, | |
| "eval_samples_per_second": 1285.631, | |
| "eval_steps_per_second": 80.546, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 22.53098009763425, | |
| "eval_loss": 5.570806503295898, | |
| "eval_runtime": 2.5775, | |
| "eval_samples_per_second": 1288.087, | |
| "eval_steps_per_second": 80.699, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 24.03304543747653, | |
| "grad_norm": 5.407034397125244, | |
| "learning_rate": 8e-05, | |
| "loss": 5.2633, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 24.03304543747653, | |
| "eval_loss": 4.99446439743042, | |
| "eval_runtime": 2.5969, | |
| "eval_samples_per_second": 1278.448, | |
| "eval_steps_per_second": 80.096, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 25.535110777318813, | |
| "eval_loss": 4.523237705230713, | |
| "eval_runtime": 2.5888, | |
| "eval_samples_per_second": 1282.462, | |
| "eval_steps_per_second": 80.347, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 27.037176117161096, | |
| "grad_norm": 5.287095069885254, | |
| "learning_rate": 9e-05, | |
| "loss": 4.3386, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 27.037176117161096, | |
| "eval_loss": 4.140077590942383, | |
| "eval_runtime": 2.5851, | |
| "eval_samples_per_second": 1284.299, | |
| "eval_steps_per_second": 80.462, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 28.539241457003378, | |
| "eval_loss": 3.857847213745117, | |
| "eval_runtime": 2.6038, | |
| "eval_samples_per_second": 1275.044, | |
| "eval_steps_per_second": 79.882, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 30.041306796845664, | |
| "grad_norm": 4.30466890335083, | |
| "learning_rate": 0.0001, | |
| "loss": 3.6823, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 30.041306796845664, | |
| "eval_loss": 3.6678993701934814, | |
| "eval_runtime": 2.7074, | |
| "eval_samples_per_second": 1226.275, | |
| "eval_steps_per_second": 76.827, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 31.543372136687946, | |
| "eval_loss": 3.5396695137023926, | |
| "eval_runtime": 4.9347, | |
| "eval_samples_per_second": 672.792, | |
| "eval_steps_per_second": 42.151, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 33.04543747653023, | |
| "grad_norm": 4.037199974060059, | |
| "learning_rate": 9.333333333333334e-05, | |
| "loss": 3.3424, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 33.04543747653023, | |
| "eval_loss": 3.4012534618377686, | |
| "eval_runtime": 4.8721, | |
| "eval_samples_per_second": 681.437, | |
| "eval_steps_per_second": 42.692, | |
| "step": 44000 | |
| } | |
| ], | |
| "logging_steps": 4000, | |
| "max_steps": 100000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 76, | |
| "save_steps": 4000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3722640439574528e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |