diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/config.json b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/config.json new file mode 100644 index 0000000000000000000000000000000000000000..19ceff9aaeb39e2e89bdc1603dd24ddc5e95c03a --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/config.json @@ -0,0 +1,69 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "clvla_vision_last_n_layers": 27, + "clvla_vision_lr_multiplier": 0.1, + "clvla_vision_total_layers": 27, + "clvla_vision_train_connector": true, + "clvla_vision_tuning_mode": "full", + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/experiment_cfg/clvla_contract.yaml b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/experiment_cfg/clvla_contract.yaml new file mode 100644 index 0000000000000000000000000000000000000000..a6f59d30ead0c655b47f5ca9a17fe91171a4fb24 --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/experiment_cfg/clvla_contract.yaml @@ -0,0 +1,51 @@ +schema_version: clvla_n15_v1 +variant: baseline +relation_mode: null +future_token_mode: checkpoint_weight +vision_tuning: + mode: full + train_connector: true + lr_multiplier: 0.1 +auxiliary: {} +relation_mismatch: {} +objective: {} +input_cache: {} +episode_shard: {} +n17_official_shard: {} +training: + dataset_path: + - /home/work/workspace/test/seonho/robocasa_mg_gr00t_300 + output_dir: /home/work/workspace/test/seonho/groot_robocasa-kitchen/train_ckpt/n15/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K + experiment_name: Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K + wandb_project: finetune-gr00t-n1d5 + data_config: single_panda_gripper + base_model_path: nvidia/GR00T-N1.5-3B + embodiment_tag: new_embodiment + video_backend: torchcodec + batch_size: 32 + max_steps: 60000 + num_gpus: 1 + save_steps: 5000 + save_total_limit: 1 + tune_llm: false + tune_visual: true + tune_projector: true + tune_diffusion_model: true + learning_rate: 0.0001 + gradient_accumulation_steps: 2 + weight_decay: 1.0e-05 + warmup_ratio: 0.05 + lora_rank: 0 + lora_alpha: 16 + lora_dropout: 0.1 + lora_full_model: false + dataloader_num_workers: 8 + dataloader_prefetch_factor: 4 + report_to: wandb + resume: false + balance_dataset_weights: true + balance_trajectory_weights: true +n15_commit: d1ccca163c96639ddc677dae3f4f14d05cbc63fd +config_path: /home/work/workspace/test/seonho/clvla/Isaac-GR00T-N1.5/clvla/cfg/vision_enc/baseline/baseline_full_vision_connector_fm_vlr0p1.yaml +config_sha256: ecfaea898d548605de762855e1451ffc35683fe8cf042cc046dcbd45db46195a +data_contract: single_panda_gripper_v1 diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/experiment_cfg/metadata.json b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..1b3e0413d6dbd29e2e26dd8593a27178687407bc --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/experiment_cfg/metadata.json @@ -0,0 +1,807 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 5.601607322692871, + -0.4872598946094513, + 0.7495372295379639 + ], + "min": [ + 0.31250813603401184, + -5.100343704223633, + 0.6963803172111511 + ], + "mean": [ + 2.5315651893615723, + -1.5020122528076172, + 0.7073471546173096 + ], + "std": [ + 1.4670116901397705, + 0.8559004664421082, + 0.006971539929509163 + ], + "q01": [ + 0.5000457167625427, + -4.100119590759277, + 0.6999446749687195 + ], + "q99": [ + 5.5999836921691895, + -0.778001925945282, + 0.7071865797042847 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.2765239179134369, + 0.6414101719856262 + ], + "std": [ + 0.0, + 0.0, + 0.6192871332168579, + 0.35383468866348267 + ], + "q01": [ + 0.0, + 0.0, + -1.0, + 5.569559434661642e-06 + ], + "q99": [ + 0.0, + 0.0, + 1.0, + 1.0 + ] + }, + "end_effector_position_absolute": { + "max": [ + 6.18405818939209, + -0.009097904898226261, + 1.8348677158355713 + ], + "min": [ + 0.01175791583955288, + -5.087834358215332, + 0.7236378788948059 + ], + "mean": [ + 2.5557069778442383, + -1.4287182092666626, + 1.2477613687515259 + ], + "std": [ + 1.626364827156067, + 0.9738650321960449, + 0.24595242738723755 + ], + "q01": [ + 0.145528781414032, + -4.12215726852417, + 0.7766175556182862 + ], + "q99": [ + 5.86918882369995, + -0.14761754989624046, + 1.6773935270309448 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.8734620809555054, + 0.8441421985626221, + 1.1348456144332886 + ], + "min": [ + -0.46594083309173584, + -0.8278615474700928, + 0.011948669329285622 + ], + "mean": [ + 0.266306072473526, + -0.04002884030342102, + 0.5537278652191162 + ], + "std": [ + 0.19139817357063293, + 0.2582354247570038, + 0.2461009919643402 + ], + "q01": [ + -0.251964715719223, + -0.6177875232696534, + 0.07541620314121246 + ], + "q99": [ + 0.736240336894989, + 0.6068667078018188, + 0.9772176861763 + ] + }, + "end_effector_rotation_absolute": { + "max": [ + 0.9999977350234985, + 0.9990691542625427, + 0.9815630316734314, + 0.9457746744155884 + ], + "min": [ + -0.9574247598648071, + -0.9999843239784241, + -0.8860878348350525, + -0.960811972618103 + ], + "mean": [ + 0.4705830514431, + 0.023915516212582588, + 0.17580851912498474, + -0.0628008246421814 + ], + "std": [ + 0.418401300907135, + 0.602892279624939, + 0.30509766936302185, + 0.33302298188209534 + ], + "q01": [ + -0.7058527708053589, + -0.9918827247619629, + -0.6562725901603699, + -0.7662235760688781 + ], + "q99": [ + 0.991655216217041, + 0.9357455682754516, + 0.7898480153083801, + 0.7428149509429931 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999980926513672, + 0.990567147731781, + 0.964046835899353, + 0.9740113019943237 + ], + "min": [ + -0.9999996423721313, + -0.9910850524902344, + -0.9571061134338379, + 6.351022108219695e-08 + ], + "mean": [ + -0.14963096380233765, + 0.1248803362250328, + 0.0027880165725946426, + 0.2082664668560028 + ], + "std": [ + 0.7683019042015076, + 0.36296239495277405, + 0.37635329365730286, + 0.2322801947593689 + ], + "q01": [ + -0.9973158955574035, + -0.6837371301651001, + -0.7814695286750794, + 0.0013687612861394884 + ], + "q99": [ + 0.9957762026786804, + 0.8491807365417481, + 0.7687161755561829, + 0.802422981262207 + ] + }, + "gripper_qpos": { + "max": [ + 0.07896614819765091, + 0.10181591659784317 + ], + "min": [ + -0.027591193094849586, + -0.05319468677043915 + ], + "mean": [ + 0.030854083597660065, + -0.03072655387222767 + ], + "std": [ + 0.013360796496272087, + 0.013391095213592052 + ], + "q01": [ + -4.0135159069905056e-05, + -0.04032127052545548 + ], + "q99": [ + 0.04066328033804893, + -0.00040660659200511887 + ] + }, + "gripper_qvel": { + "max": [ + 1.6098369359970093, + 16.347944259643555 + ], + "min": [ + -2.2620184421539307, + -0.607681393623352 + ], + "mean": [ + 0.000719624396879226, + -0.0007039760239422321 + ], + "std": [ + 0.014941745437681675, + 0.01924414187669754 + ], + "q01": [ + -0.06061995089054108, + -0.061852154731750486 + ], + "q99": [ + 0.060133510529994924, + 0.06022836208343506 + ] + }, + "joint_position": { + "max": [ + 2.9067223072052, + 1.4987837076187134, + 2.899402141571045, + -0.02150430530309677, + 2.9257497787475586, + 3.788130760192871, + 2.9191486835479736 + ], + "min": [ + -2.9045588970184326, + -1.778917908668518, + -2.898763656616211, + -3.0899336338043213, + -2.919085741043091, + 0.03439585119485855, + -2.9150876998901367 + ], + "mean": [ + -0.009502203203737736, + -0.8110345005989075, + -0.14543816447257996, + -2.092560291290283, + 0.21085135638713837, + 1.9832311868667603, + 0.6437883973121643 + ], + "std": [ + 0.3928738534450531, + 0.615814745426178, + 0.6957448124885559, + 0.6573575139045715, + 0.7360686659812927, + 0.5835347771644592, + 0.8847095966339111 + ], + "q01": [ + -1.112356686592102, + -1.7629162073135376, + -1.952404227256775, + -3.0719614028930664, + -1.3566890811920167, + 0.713577287197113, + -1.6071688652038574 + ], + "q99": [ + 1.4290750741958596, + 0.7505003905296316, + 1.8789696216583252, + -0.06965330421924591, + 2.897454261779785, + 3.5309412384033183, + 2.7498557090759275 + ] + }, + "joint_position_cos": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.9997687935829163, + 1.0, + 0.9994085431098938, + 1.0 + ], + "min": [ + -0.9725445508956909, + -0.20662237703800201, + -0.9708149433135986, + -0.998665988445282, + -0.9767962098121643, + -1.0, + -0.975361168384552 + ], + "mean": [ + 0.9368533492088318, + 0.5487301349639893, + 0.7909337282180786, + -0.44477057456970215, + 0.8003951907157898, + -0.3284209966659546, + 0.5292185544967651 + ], + "std": [ + 0.18806983530521393, + 0.3238338530063629, + 0.3483835756778717, + 0.5128603577613831, + 0.41907837986946106, + 0.44435837864875793, + 0.4577120542526245 + ], + "q01": [ + -0.028352123275399165, + -0.19094021618366241, + -0.4218991422653198, + -0.9975767135620117, + -0.9703471064567566, + -0.9987616801261902, + -0.9365046453475953 + ], + "q99": [ + 0.9999994039535522, + 0.9995378518104553, + 0.9999944567680359, + 0.9975751638412476, + 0.9999942183494568, + 0.7560252261161804, + 0.9997037053108215 + ] + }, + "joint_position_sin": { + "max": [ + 1.0, + 0.9974082112312317, + 1.0, + -0.021502649411559105, + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + -1.0, + -1.0, + -0.6024267673492432, + -1.0 + ], + "mean": [ + -0.016687549650669098, + -0.6223337650299072, + -0.1208081841468811, + -0.6835928559303284, + 0.10104460269212723, + 0.7790007591247559, + 0.4382820129394531 + ], + "std": [ + 0.3047696352005005, + 0.4566800892353058, + 0.49255555868148804, + 0.26993513107299805, + 0.4195447862148285, + 0.30101412534713745, + 0.5644168257713318 + ], + "q01": [ + -0.8887296628952026, + -0.9998934864997864, + -0.9991217708587646, + -0.9997780919075012, + -0.9557623648643494, + -0.3795857656002044, + -0.987897298336029 + ], + "q99": [ + 0.9497140288352965, + 0.6820047998428338, + 0.9963063192367554, + -0.06942577660083771, + 0.9981896734237671, + 0.9999741911888123, + 0.9997917675971985 + ] + }, + "joint_velocity": { + "max": [ + 3.9610798358917236, + 4.708287715911865, + 4.893841743469238, + 6.32977819442749, + 9.424575805664062, + 8.002043724060059, + 6.2579474449157715 + ], + "min": [ + -3.6412594318389893, + -3.3275678157806396, + -4.578517436981201, + -5.057194232940674, + -6.067282199859619, + -6.464511871337891, + -6.37498664855957 + ], + "mean": [ + -0.004871963057667017, + 0.017353052273392677, + -0.0031993486918509007, + 0.018676184117794037, + 0.011898555792868137, + 0.03337086737155914, + -0.005988204386085272 + ], + "std": [ + 0.22286543250083923, + 0.3334001302719116, + 0.2544739544391632, + 0.36149299144744873, + 0.33571285009384155, + 0.38648521900177, + 0.3577897250652313 + ], + "q01": [ + -0.7382641816139222, + -0.9932713437080383, + -0.6891296458244324, + -1.0751154088973998, + -0.9251720833778382, + -1.05612979888916, + -1.0389645385742188 + ], + "q99": [ + 0.6494716310501091, + 0.8279203224182128, + 0.7093800806999206, + 1.0475247430801367, + 1.059649910926817, + 1.056202611923216, + 0.9718989586830125 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 0.0, + 0.0, + 0.0, + 0.0 + ], + "min": [ + 0.0, + 0.0, + 0.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.0, + 0.0 + ], + "std": [ + 0.0, + 0.0, + 0.0, + 0.0 + ], + "q01": [ + 0.0, + 0.0, + 0.0, + 0.0 + ], + "q99": [ + 0.0, + 0.0, + 0.0, + 0.0 + ] + }, + "control_mode": { + "max": [ + 0.0 + ], + "min": [ + 0.0 + ], + "mean": [ + 0.0 + ], + "std": [ + 0.0 + ], + "q01": [ + 0.0 + ], + "q99": [ + 0.0 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.029150759801268578, + 0.007607606239616871, + -0.037477411329746246 + ], + "std": [ + 0.4175023138523102, + 0.41343387961387634, + 0.36864909529685974 + ], + "q01": [ + -1.0, + -1.0, + -1.0 + ], + "q99": [ + 1.0, + 1.0, + 1.0 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.007235987111926079, + -0.023321982473134995, + -0.0024119403678923845 + ], + "std": [ + 0.147710919380188, + 0.15399158000946045, + 0.13929523527622223 + ], + "q01": [ + -0.42321433901786804, + -0.5219977021217346, + -0.4101182746887207 + ], + "q99": [ + 0.47503985762596074, + 0.38744416117668146, + 0.4242807936668394 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + 0.0 + ], + "mean": [ + 0.34521719813346863 + ], + "std": [ + 0.47949448227882385 + ], + "q01": [ + 0.0 + ], + "q99": [ + 1.0 + ] + } + } + }, + "modalities": { + "video": { + "left_view": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "right_view": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "wrist_view": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_absolute": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_absolute": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + }, + "gripper_qvel": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + }, + "joint_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 7 + ], + "continuous": true + }, + "joint_position_cos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 7 + ], + "continuous": true + }, + "joint_position_sin": { + "absolute": true, + "rotation_type": null, + "shape": [ + 7 + ], + "continuous": true + }, + "joint_velocity": { + "absolute": true, + "rotation_type": null, + "shape": [ + 7 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": false, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": false + }, + "end_effector_position": { + "absolute": false, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": false, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": false + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model-00001-of-00002.safetensors b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b35d6013105583e06abe013cdf4259d79eba9e2b --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9347848c6f9af894f4f8054c2c94cb032cc08dbf99dc6563f588c6459452722 +size 4999367032 diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model-00002-of-00002.safetensors b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..480f49c1bdfa443fcd4556258410ed9893b1ccf8 --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59fab8dfbd19a850de08e223d85c3ec44413b19e55b0d86f12d0f50f8480d7df +size 2586705312 diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model.safetensors.index.json b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/trainer_state.json b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..a6ea762a24844581ae492ff27fea3779e18d9aa3 --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/trainer_state.json @@ -0,0 +1,42043 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.851966170751281, + "eval_steps": 500, + "global_step": 60000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00030866102845854683, + "grad_norm": 6.432484149932861, + "learning_rate": 3.0000000000000004e-07, + "loss": 0.5599, + "step": 10 + }, + { + "epoch": 0.0006173220569170937, + "grad_norm": 5.993962287902832, + "learning_rate": 6.333333333333333e-07, + "loss": 0.5452, + "step": 20 + }, + { + "epoch": 0.0009259830853756405, + "grad_norm": 6.0441060066223145, + "learning_rate": 9.666666666666668e-07, + "loss": 0.5124, + "step": 30 + }, + { + "epoch": 0.0012346441138341873, + "grad_norm": 4.536877632141113, + "learning_rate": 1.3e-06, + "loss": 0.4136, + "step": 40 + }, + { + "epoch": 0.001543305142292734, + "grad_norm": 1.96611487865448, + "learning_rate": 1.6333333333333333e-06, + "loss": 0.2944, + "step": 50 + }, + { + "epoch": 0.001851966170751281, + "grad_norm": 1.7445751428604126, + "learning_rate": 1.9666666666666668e-06, + "loss": 0.2375, + "step": 60 + }, + { + "epoch": 0.002160627199209828, + "grad_norm": 1.3558547496795654, + "learning_rate": 2.3e-06, + "loss": 0.22, + "step": 70 + }, + { + "epoch": 0.0024692882276683746, + "grad_norm": 0.8783726692199707, + "learning_rate": 2.6333333333333337e-06, + "loss": 0.1778, + "step": 80 + }, + { + "epoch": 0.0027779492561269214, + "grad_norm": 0.7693445682525635, + "learning_rate": 2.966666666666667e-06, + "loss": 0.1614, + "step": 90 + }, + { + "epoch": 0.003086610284585468, + "grad_norm": 0.8126460313796997, + "learning_rate": 3.3e-06, + "loss": 0.1487, + "step": 100 + }, + { + "epoch": 0.0033952713130440152, + "grad_norm": 0.521102249622345, + "learning_rate": 3.633333333333334e-06, + "loss": 0.1501, + "step": 110 + }, + { + "epoch": 0.003703932341502562, + "grad_norm": 0.6465502381324768, + "learning_rate": 3.966666666666667e-06, + "loss": 0.1262, + "step": 120 + }, + { + "epoch": 0.004012593369961109, + "grad_norm": 0.5397230386734009, + "learning_rate": 4.2999999999999995e-06, + "loss": 0.1247, + "step": 130 + }, + { + "epoch": 0.004321254398419656, + "grad_norm": 0.6289680600166321, + "learning_rate": 4.633333333333334e-06, + "loss": 0.1335, + "step": 140 + }, + { + "epoch": 0.004629915426878202, + "grad_norm": 0.4911990463733673, + "learning_rate": 4.966666666666667e-06, + "loss": 0.1359, + "step": 150 + }, + { + "epoch": 0.004938576455336749, + "grad_norm": 0.4905587434768677, + "learning_rate": 5.3e-06, + "loss": 0.1282, + "step": 160 + }, + { + "epoch": 0.005247237483795296, + "grad_norm": 0.6799914240837097, + "learning_rate": 5.633333333333333e-06, + "loss": 0.1193, + "step": 170 + }, + { + "epoch": 0.005555898512253843, + "grad_norm": 0.4684731960296631, + "learning_rate": 5.9666666666666666e-06, + "loss": 0.1113, + "step": 180 + }, + { + "epoch": 0.00586455954071239, + "grad_norm": 0.6558704972267151, + "learning_rate": 6.300000000000001e-06, + "loss": 0.1153, + "step": 190 + }, + { + "epoch": 0.006173220569170936, + "grad_norm": 0.5794803500175476, + "learning_rate": 6.633333333333333e-06, + "loss": 0.1147, + "step": 200 + }, + { + "epoch": 0.006481881597629483, + "grad_norm": 0.8498133420944214, + "learning_rate": 6.966666666666667e-06, + "loss": 0.1156, + "step": 210 + }, + { + "epoch": 0.0067905426260880305, + "grad_norm": 0.6754714250564575, + "learning_rate": 7.2999999999999996e-06, + "loss": 0.1123, + "step": 220 + }, + { + "epoch": 0.007099203654546577, + "grad_norm": 0.6554245352745056, + "learning_rate": 7.633333333333334e-06, + "loss": 0.1166, + "step": 230 + }, + { + "epoch": 0.007407864683005124, + "grad_norm": 0.6282513737678528, + "learning_rate": 7.966666666666666e-06, + "loss": 0.1228, + "step": 240 + }, + { + "epoch": 0.00771652571146367, + "grad_norm": 0.6241544485092163, + "learning_rate": 8.3e-06, + "loss": 0.1101, + "step": 250 + }, + { + "epoch": 0.008025186739922217, + "grad_norm": 0.4563302993774414, + "learning_rate": 8.633333333333334e-06, + "loss": 0.1043, + "step": 260 + }, + { + "epoch": 0.008333847768380764, + "grad_norm": 0.4674703776836395, + "learning_rate": 8.966666666666668e-06, + "loss": 0.1059, + "step": 270 + }, + { + "epoch": 0.008642508796839312, + "grad_norm": 0.5488001704216003, + "learning_rate": 9.3e-06, + "loss": 0.1057, + "step": 280 + }, + { + "epoch": 0.008951169825297858, + "grad_norm": 0.41368240118026733, + "learning_rate": 9.633333333333335e-06, + "loss": 0.0948, + "step": 290 + }, + { + "epoch": 0.009259830853756404, + "grad_norm": 0.8926979303359985, + "learning_rate": 9.966666666666667e-06, + "loss": 0.1102, + "step": 300 + }, + { + "epoch": 0.009568491882214952, + "grad_norm": 0.7092853784561157, + "learning_rate": 1.03e-05, + "loss": 0.107, + "step": 310 + }, + { + "epoch": 0.009877152910673499, + "grad_norm": 0.48313990235328674, + "learning_rate": 1.0633333333333334e-05, + "loss": 0.0949, + "step": 320 + }, + { + "epoch": 0.010185813939132045, + "grad_norm": 0.5668746829032898, + "learning_rate": 1.0966666666666666e-05, + "loss": 0.0984, + "step": 330 + }, + { + "epoch": 0.010494474967590591, + "grad_norm": 0.6383707523345947, + "learning_rate": 1.13e-05, + "loss": 0.1103, + "step": 340 + }, + { + "epoch": 0.01080313599604914, + "grad_norm": 0.6056993007659912, + "learning_rate": 1.1633333333333334e-05, + "loss": 0.0969, + "step": 350 + }, + { + "epoch": 0.011111797024507685, + "grad_norm": 0.7079081535339355, + "learning_rate": 1.1966666666666668e-05, + "loss": 0.099, + "step": 360 + }, + { + "epoch": 0.011420458052966232, + "grad_norm": 0.6560660600662231, + "learning_rate": 1.23e-05, + "loss": 0.0924, + "step": 370 + }, + { + "epoch": 0.01172911908142478, + "grad_norm": 0.8999735713005066, + "learning_rate": 1.2633333333333333e-05, + "loss": 0.0877, + "step": 380 + }, + { + "epoch": 0.012037780109883326, + "grad_norm": 0.539846658706665, + "learning_rate": 1.2966666666666669e-05, + "loss": 0.0868, + "step": 390 + }, + { + "epoch": 0.012346441138341872, + "grad_norm": 0.5094746351242065, + "learning_rate": 1.3300000000000001e-05, + "loss": 0.0892, + "step": 400 + }, + { + "epoch": 0.01265510216680042, + "grad_norm": 0.5733831524848938, + "learning_rate": 1.3633333333333334e-05, + "loss": 0.0997, + "step": 410 + }, + { + "epoch": 0.012963763195258967, + "grad_norm": 0.5628836154937744, + "learning_rate": 1.3966666666666666e-05, + "loss": 0.0896, + "step": 420 + }, + { + "epoch": 0.013272424223717513, + "grad_norm": 0.6709559559822083, + "learning_rate": 1.43e-05, + "loss": 0.102, + "step": 430 + }, + { + "epoch": 0.013581085252176061, + "grad_norm": 0.5378535389900208, + "learning_rate": 1.4633333333333334e-05, + "loss": 0.0919, + "step": 440 + }, + { + "epoch": 0.013889746280634607, + "grad_norm": 0.47225749492645264, + "learning_rate": 1.4966666666666668e-05, + "loss": 0.089, + "step": 450 + }, + { + "epoch": 0.014198407309093154, + "grad_norm": 0.8069123029708862, + "learning_rate": 1.53e-05, + "loss": 0.0889, + "step": 460 + }, + { + "epoch": 0.014507068337551702, + "grad_norm": 0.5842798948287964, + "learning_rate": 1.563333333333333e-05, + "loss": 0.0858, + "step": 470 + }, + { + "epoch": 0.014815729366010248, + "grad_norm": 0.5566721558570862, + "learning_rate": 1.5966666666666667e-05, + "loss": 0.0911, + "step": 480 + }, + { + "epoch": 0.015124390394468794, + "grad_norm": 0.6334349513053894, + "learning_rate": 1.63e-05, + "loss": 0.0901, + "step": 490 + }, + { + "epoch": 0.01543305142292734, + "grad_norm": 0.47930827736854553, + "learning_rate": 1.6633333333333336e-05, + "loss": 0.0906, + "step": 500 + }, + { + "epoch": 0.01574171245138589, + "grad_norm": 0.4856976568698883, + "learning_rate": 1.6966666666666668e-05, + "loss": 0.0889, + "step": 510 + }, + { + "epoch": 0.016050373479844435, + "grad_norm": 0.7145525813102722, + "learning_rate": 1.73e-05, + "loss": 0.0887, + "step": 520 + }, + { + "epoch": 0.01635903450830298, + "grad_norm": 0.5659053921699524, + "learning_rate": 1.7633333333333336e-05, + "loss": 0.0908, + "step": 530 + }, + { + "epoch": 0.016667695536761527, + "grad_norm": 0.6952089667320251, + "learning_rate": 1.796666666666667e-05, + "loss": 0.0946, + "step": 540 + }, + { + "epoch": 0.016976356565220074, + "grad_norm": 0.5707944631576538, + "learning_rate": 1.83e-05, + "loss": 0.0802, + "step": 550 + }, + { + "epoch": 0.017285017593678623, + "grad_norm": 0.83467698097229, + "learning_rate": 1.8633333333333333e-05, + "loss": 0.0888, + "step": 560 + }, + { + "epoch": 0.01759367862213717, + "grad_norm": 0.5621892213821411, + "learning_rate": 1.896666666666667e-05, + "loss": 0.093, + "step": 570 + }, + { + "epoch": 0.017902339650595716, + "grad_norm": 0.6784348487854004, + "learning_rate": 1.93e-05, + "loss": 0.0918, + "step": 580 + }, + { + "epoch": 0.018211000679054262, + "grad_norm": 0.6393017172813416, + "learning_rate": 1.9633333333333334e-05, + "loss": 0.0836, + "step": 590 + }, + { + "epoch": 0.01851966170751281, + "grad_norm": 0.5687531232833862, + "learning_rate": 1.9966666666666666e-05, + "loss": 0.0934, + "step": 600 + }, + { + "epoch": 0.018828322735971355, + "grad_norm": 0.548298716545105, + "learning_rate": 2.0300000000000002e-05, + "loss": 0.0853, + "step": 610 + }, + { + "epoch": 0.019136983764429905, + "grad_norm": 0.6980525851249695, + "learning_rate": 2.0633333333333335e-05, + "loss": 0.0854, + "step": 620 + }, + { + "epoch": 0.01944564479288845, + "grad_norm": 0.49039676785469055, + "learning_rate": 2.0966666666666667e-05, + "loss": 0.0861, + "step": 630 + }, + { + "epoch": 0.019754305821346997, + "grad_norm": 0.5117352604866028, + "learning_rate": 2.13e-05, + "loss": 0.0895, + "step": 640 + }, + { + "epoch": 0.020062966849805543, + "grad_norm": 0.4894750118255615, + "learning_rate": 2.1633333333333332e-05, + "loss": 0.0875, + "step": 650 + }, + { + "epoch": 0.02037162787826409, + "grad_norm": 0.5976028442382812, + "learning_rate": 2.1966666666666668e-05, + "loss": 0.0841, + "step": 660 + }, + { + "epoch": 0.020680288906722636, + "grad_norm": 0.6428791880607605, + "learning_rate": 2.23e-05, + "loss": 0.0836, + "step": 670 + }, + { + "epoch": 0.020988949935181182, + "grad_norm": 0.4743877947330475, + "learning_rate": 2.2633333333333336e-05, + "loss": 0.0817, + "step": 680 + }, + { + "epoch": 0.021297610963639732, + "grad_norm": 0.7302097082138062, + "learning_rate": 2.2966666666666668e-05, + "loss": 0.0794, + "step": 690 + }, + { + "epoch": 0.02160627199209828, + "grad_norm": 0.6103028059005737, + "learning_rate": 2.3300000000000004e-05, + "loss": 0.0831, + "step": 700 + }, + { + "epoch": 0.021914933020556825, + "grad_norm": 0.761784017086029, + "learning_rate": 2.3633333333333336e-05, + "loss": 0.0843, + "step": 710 + }, + { + "epoch": 0.02222359404901537, + "grad_norm": 0.5122126340866089, + "learning_rate": 2.396666666666667e-05, + "loss": 0.0817, + "step": 720 + }, + { + "epoch": 0.022532255077473917, + "grad_norm": 0.6548123955726624, + "learning_rate": 2.43e-05, + "loss": 0.0798, + "step": 730 + }, + { + "epoch": 0.022840916105932463, + "grad_norm": 0.6320444345474243, + "learning_rate": 2.4633333333333334e-05, + "loss": 0.0847, + "step": 740 + }, + { + "epoch": 0.023149577134391013, + "grad_norm": 0.4701177477836609, + "learning_rate": 2.496666666666667e-05, + "loss": 0.0839, + "step": 750 + }, + { + "epoch": 0.02345823816284956, + "grad_norm": 0.4762485921382904, + "learning_rate": 2.5300000000000002e-05, + "loss": 0.0892, + "step": 760 + }, + { + "epoch": 0.023766899191308106, + "grad_norm": 0.6298289895057678, + "learning_rate": 2.5633333333333338e-05, + "loss": 0.0869, + "step": 770 + }, + { + "epoch": 0.024075560219766652, + "grad_norm": 0.5555379986763, + "learning_rate": 2.5966666666666667e-05, + "loss": 0.0824, + "step": 780 + }, + { + "epoch": 0.0243842212482252, + "grad_norm": 0.4853808581829071, + "learning_rate": 2.6300000000000002e-05, + "loss": 0.0832, + "step": 790 + }, + { + "epoch": 0.024692882276683745, + "grad_norm": 0.5520538091659546, + "learning_rate": 2.663333333333333e-05, + "loss": 0.0803, + "step": 800 + }, + { + "epoch": 0.025001543305142294, + "grad_norm": 0.41240960359573364, + "learning_rate": 2.6966666666666667e-05, + "loss": 0.0707, + "step": 810 + }, + { + "epoch": 0.02531020433360084, + "grad_norm": 0.537031888961792, + "learning_rate": 2.7300000000000003e-05, + "loss": 0.0775, + "step": 820 + }, + { + "epoch": 0.025618865362059387, + "grad_norm": 0.5477346777915955, + "learning_rate": 2.7633333333333332e-05, + "loss": 0.0805, + "step": 830 + }, + { + "epoch": 0.025927526390517933, + "grad_norm": 0.8339902758598328, + "learning_rate": 2.7966666666666668e-05, + "loss": 0.0842, + "step": 840 + }, + { + "epoch": 0.02623618741897648, + "grad_norm": 0.46177244186401367, + "learning_rate": 2.83e-05, + "loss": 0.0806, + "step": 850 + }, + { + "epoch": 0.026544848447435026, + "grad_norm": 0.6566048264503479, + "learning_rate": 2.8633333333333336e-05, + "loss": 0.0844, + "step": 860 + }, + { + "epoch": 0.026853509475893572, + "grad_norm": 0.47063419222831726, + "learning_rate": 2.8966666666666668e-05, + "loss": 0.0786, + "step": 870 + }, + { + "epoch": 0.027162170504352122, + "grad_norm": 0.486644446849823, + "learning_rate": 2.93e-05, + "loss": 0.0825, + "step": 880 + }, + { + "epoch": 0.027470831532810668, + "grad_norm": 0.6081055402755737, + "learning_rate": 2.9633333333333336e-05, + "loss": 0.0817, + "step": 890 + }, + { + "epoch": 0.027779492561269215, + "grad_norm": 0.4567124843597412, + "learning_rate": 2.9966666666666672e-05, + "loss": 0.0825, + "step": 900 + }, + { + "epoch": 0.02808815358972776, + "grad_norm": 0.4813464283943176, + "learning_rate": 3.03e-05, + "loss": 0.0765, + "step": 910 + }, + { + "epoch": 0.028396814618186307, + "grad_norm": 0.5592803955078125, + "learning_rate": 3.063333333333334e-05, + "loss": 0.0749, + "step": 920 + }, + { + "epoch": 0.028705475646644853, + "grad_norm": 0.4505971670150757, + "learning_rate": 3.096666666666666e-05, + "loss": 0.0767, + "step": 930 + }, + { + "epoch": 0.029014136675103403, + "grad_norm": 0.47470393776893616, + "learning_rate": 3.13e-05, + "loss": 0.0777, + "step": 940 + }, + { + "epoch": 0.02932279770356195, + "grad_norm": 0.4791828691959381, + "learning_rate": 3.1633333333333334e-05, + "loss": 0.0763, + "step": 950 + }, + { + "epoch": 0.029631458732020496, + "grad_norm": 0.661306619644165, + "learning_rate": 3.196666666666667e-05, + "loss": 0.0815, + "step": 960 + }, + { + "epoch": 0.029940119760479042, + "grad_norm": 0.5632457733154297, + "learning_rate": 3.2300000000000006e-05, + "loss": 0.0805, + "step": 970 + }, + { + "epoch": 0.03024878078893759, + "grad_norm": 0.5635040998458862, + "learning_rate": 3.263333333333333e-05, + "loss": 0.076, + "step": 980 + }, + { + "epoch": 0.030557441817396135, + "grad_norm": 0.5389256477355957, + "learning_rate": 3.296666666666667e-05, + "loss": 0.0891, + "step": 990 + }, + { + "epoch": 0.03086610284585468, + "grad_norm": 0.6321320533752441, + "learning_rate": 3.33e-05, + "loss": 0.0755, + "step": 1000 + }, + { + "epoch": 0.03117476387431323, + "grad_norm": 0.4636113941669464, + "learning_rate": 3.3633333333333335e-05, + "loss": 0.0714, + "step": 1010 + }, + { + "epoch": 0.03148342490277178, + "grad_norm": 0.4456343650817871, + "learning_rate": 3.396666666666667e-05, + "loss": 0.0807, + "step": 1020 + }, + { + "epoch": 0.03179208593123032, + "grad_norm": 0.6430743932723999, + "learning_rate": 3.430000000000001e-05, + "loss": 0.0768, + "step": 1030 + }, + { + "epoch": 0.03210074695968887, + "grad_norm": 0.5543654561042786, + "learning_rate": 3.463333333333333e-05, + "loss": 0.0752, + "step": 1040 + }, + { + "epoch": 0.03240940798814742, + "grad_norm": 0.6094911694526672, + "learning_rate": 3.496666666666667e-05, + "loss": 0.0751, + "step": 1050 + }, + { + "epoch": 0.03271806901660596, + "grad_norm": 0.4230935275554657, + "learning_rate": 3.53e-05, + "loss": 0.0725, + "step": 1060 + }, + { + "epoch": 0.03302673004506451, + "grad_norm": 0.5054906606674194, + "learning_rate": 3.563333333333334e-05, + "loss": 0.0764, + "step": 1070 + }, + { + "epoch": 0.033335391073523055, + "grad_norm": 0.37808412313461304, + "learning_rate": 3.596666666666667e-05, + "loss": 0.0802, + "step": 1080 + }, + { + "epoch": 0.033644052101981604, + "grad_norm": 0.4492661952972412, + "learning_rate": 3.63e-05, + "loss": 0.0751, + "step": 1090 + }, + { + "epoch": 0.03395271313044015, + "grad_norm": 0.5126152038574219, + "learning_rate": 3.6633333333333334e-05, + "loss": 0.0711, + "step": 1100 + }, + { + "epoch": 0.0342613741588987, + "grad_norm": 0.5184963345527649, + "learning_rate": 3.6966666666666666e-05, + "loss": 0.0719, + "step": 1110 + }, + { + "epoch": 0.03457003518735725, + "grad_norm": 0.4447159469127655, + "learning_rate": 3.73e-05, + "loss": 0.0722, + "step": 1120 + }, + { + "epoch": 0.03487869621581579, + "grad_norm": 0.4048321545124054, + "learning_rate": 3.763333333333334e-05, + "loss": 0.0793, + "step": 1130 + }, + { + "epoch": 0.03518735724427434, + "grad_norm": 0.43180492520332336, + "learning_rate": 3.796666666666667e-05, + "loss": 0.0756, + "step": 1140 + }, + { + "epoch": 0.03549601827273288, + "grad_norm": 0.43074554204940796, + "learning_rate": 3.83e-05, + "loss": 0.0767, + "step": 1150 + }, + { + "epoch": 0.03580467930119143, + "grad_norm": 0.3747962415218353, + "learning_rate": 3.8633333333333335e-05, + "loss": 0.0754, + "step": 1160 + }, + { + "epoch": 0.03611334032964998, + "grad_norm": 0.4327808916568756, + "learning_rate": 3.896666666666667e-05, + "loss": 0.0759, + "step": 1170 + }, + { + "epoch": 0.036422001358108524, + "grad_norm": 0.5380723476409912, + "learning_rate": 3.9300000000000007e-05, + "loss": 0.0756, + "step": 1180 + }, + { + "epoch": 0.036730662386567074, + "grad_norm": 0.4332505762577057, + "learning_rate": 3.963333333333333e-05, + "loss": 0.0722, + "step": 1190 + }, + { + "epoch": 0.03703932341502562, + "grad_norm": 0.6289923191070557, + "learning_rate": 3.996666666666667e-05, + "loss": 0.0783, + "step": 1200 + }, + { + "epoch": 0.03734798444348417, + "grad_norm": 0.4450661242008209, + "learning_rate": 4.0300000000000004e-05, + "loss": 0.0719, + "step": 1210 + }, + { + "epoch": 0.03765664547194271, + "grad_norm": 0.49721360206604004, + "learning_rate": 4.0633333333333336e-05, + "loss": 0.0735, + "step": 1220 + }, + { + "epoch": 0.03796530650040126, + "grad_norm": 0.4662691652774811, + "learning_rate": 4.096666666666667e-05, + "loss": 0.0744, + "step": 1230 + }, + { + "epoch": 0.03827396752885981, + "grad_norm": 0.39091771841049194, + "learning_rate": 4.13e-05, + "loss": 0.074, + "step": 1240 + }, + { + "epoch": 0.03858262855731835, + "grad_norm": 0.3657782971858978, + "learning_rate": 4.1633333333333333e-05, + "loss": 0.0706, + "step": 1250 + }, + { + "epoch": 0.0388912895857769, + "grad_norm": 0.5574886798858643, + "learning_rate": 4.196666666666667e-05, + "loss": 0.0758, + "step": 1260 + }, + { + "epoch": 0.039199950614235445, + "grad_norm": 0.4543752670288086, + "learning_rate": 4.23e-05, + "loss": 0.0718, + "step": 1270 + }, + { + "epoch": 0.039508611642693994, + "grad_norm": 0.6002814769744873, + "learning_rate": 4.263333333333334e-05, + "loss": 0.0767, + "step": 1280 + }, + { + "epoch": 0.03981727267115254, + "grad_norm": 0.4292360842227936, + "learning_rate": 4.296666666666666e-05, + "loss": 0.0768, + "step": 1290 + }, + { + "epoch": 0.04012593369961109, + "grad_norm": 0.429066926240921, + "learning_rate": 4.33e-05, + "loss": 0.0769, + "step": 1300 + }, + { + "epoch": 0.04043459472806964, + "grad_norm": 0.6445325016975403, + "learning_rate": 4.3633333333333335e-05, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.04074325575652818, + "grad_norm": 0.40785253047943115, + "learning_rate": 4.396666666666667e-05, + "loss": 0.0687, + "step": 1320 + }, + { + "epoch": 0.04105191678498673, + "grad_norm": 0.5036250352859497, + "learning_rate": 4.43e-05, + "loss": 0.0774, + "step": 1330 + }, + { + "epoch": 0.04136057781344527, + "grad_norm": 0.5388437509536743, + "learning_rate": 4.463333333333334e-05, + "loss": 0.0756, + "step": 1340 + }, + { + "epoch": 0.04166923884190382, + "grad_norm": 0.5469610691070557, + "learning_rate": 4.496666666666667e-05, + "loss": 0.0784, + "step": 1350 + }, + { + "epoch": 0.041977899870362365, + "grad_norm": 0.4528872072696686, + "learning_rate": 4.53e-05, + "loss": 0.0725, + "step": 1360 + }, + { + "epoch": 0.042286560898820914, + "grad_norm": 0.4649236500263214, + "learning_rate": 4.5633333333333336e-05, + "loss": 0.0716, + "step": 1370 + }, + { + "epoch": 0.042595221927279464, + "grad_norm": 0.45635029673576355, + "learning_rate": 4.596666666666667e-05, + "loss": 0.068, + "step": 1380 + }, + { + "epoch": 0.04290388295573801, + "grad_norm": 0.4460427761077881, + "learning_rate": 4.630000000000001e-05, + "loss": 0.074, + "step": 1390 + }, + { + "epoch": 0.04321254398419656, + "grad_norm": 0.5132517218589783, + "learning_rate": 4.663333333333333e-05, + "loss": 0.0753, + "step": 1400 + }, + { + "epoch": 0.0435212050126551, + "grad_norm": 0.6834095120429993, + "learning_rate": 4.696666666666667e-05, + "loss": 0.0727, + "step": 1410 + }, + { + "epoch": 0.04382986604111365, + "grad_norm": 0.5975807905197144, + "learning_rate": 4.73e-05, + "loss": 0.0738, + "step": 1420 + }, + { + "epoch": 0.0441385270695722, + "grad_norm": 0.42646485567092896, + "learning_rate": 4.763333333333334e-05, + "loss": 0.0681, + "step": 1430 + }, + { + "epoch": 0.04444718809803074, + "grad_norm": 0.41347619891166687, + "learning_rate": 4.796666666666667e-05, + "loss": 0.0663, + "step": 1440 + }, + { + "epoch": 0.04475584912648929, + "grad_norm": 0.37798112630844116, + "learning_rate": 4.83e-05, + "loss": 0.073, + "step": 1450 + }, + { + "epoch": 0.045064510154947834, + "grad_norm": 0.5435481071472168, + "learning_rate": 4.8633333333333334e-05, + "loss": 0.0743, + "step": 1460 + }, + { + "epoch": 0.045373171183406384, + "grad_norm": 0.43607595562934875, + "learning_rate": 4.8966666666666667e-05, + "loss": 0.0701, + "step": 1470 + }, + { + "epoch": 0.04568183221186493, + "grad_norm": 0.399002343416214, + "learning_rate": 4.93e-05, + "loss": 0.0716, + "step": 1480 + }, + { + "epoch": 0.04599049324032348, + "grad_norm": 0.5146471261978149, + "learning_rate": 4.963333333333334e-05, + "loss": 0.0743, + "step": 1490 + }, + { + "epoch": 0.046299154268782027, + "grad_norm": 0.3523149788379669, + "learning_rate": 4.996666666666667e-05, + "loss": 0.0722, + "step": 1500 + }, + { + "epoch": 0.04660781529724057, + "grad_norm": 0.7448615431785583, + "learning_rate": 5.03e-05, + "loss": 0.0764, + "step": 1510 + }, + { + "epoch": 0.04691647632569912, + "grad_norm": 0.3283931314945221, + "learning_rate": 5.0633333333333335e-05, + "loss": 0.064, + "step": 1520 + }, + { + "epoch": 0.04722513735415766, + "grad_norm": 0.5337389707565308, + "learning_rate": 5.0966666666666674e-05, + "loss": 0.0756, + "step": 1530 + }, + { + "epoch": 0.04753379838261621, + "grad_norm": 0.5408918261528015, + "learning_rate": 5.130000000000001e-05, + "loss": 0.0743, + "step": 1540 + }, + { + "epoch": 0.047842459411074754, + "grad_norm": 0.3724633753299713, + "learning_rate": 5.163333333333333e-05, + "loss": 0.074, + "step": 1550 + }, + { + "epoch": 0.048151120439533304, + "grad_norm": 0.3145769238471985, + "learning_rate": 5.196666666666667e-05, + "loss": 0.0728, + "step": 1560 + }, + { + "epoch": 0.048459781467991854, + "grad_norm": 0.4329846203327179, + "learning_rate": 5.2300000000000004e-05, + "loss": 0.074, + "step": 1570 + }, + { + "epoch": 0.0487684424964504, + "grad_norm": 0.32479920983314514, + "learning_rate": 5.2633333333333336e-05, + "loss": 0.0734, + "step": 1580 + }, + { + "epoch": 0.04907710352490895, + "grad_norm": 0.5839951038360596, + "learning_rate": 5.296666666666666e-05, + "loss": 0.0711, + "step": 1590 + }, + { + "epoch": 0.04938576455336749, + "grad_norm": 0.3755994141101837, + "learning_rate": 5.330000000000001e-05, + "loss": 0.0759, + "step": 1600 + }, + { + "epoch": 0.04969442558182604, + "grad_norm": 0.3999602794647217, + "learning_rate": 5.3633333333333334e-05, + "loss": 0.0721, + "step": 1610 + }, + { + "epoch": 0.05000308661028459, + "grad_norm": 0.4395827651023865, + "learning_rate": 5.3966666666666666e-05, + "loss": 0.0678, + "step": 1620 + }, + { + "epoch": 0.05031174763874313, + "grad_norm": 0.3577362298965454, + "learning_rate": 5.4300000000000005e-05, + "loss": 0.0708, + "step": 1630 + }, + { + "epoch": 0.05062040866720168, + "grad_norm": 0.4398111402988434, + "learning_rate": 5.463333333333334e-05, + "loss": 0.0715, + "step": 1640 + }, + { + "epoch": 0.050929069695660224, + "grad_norm": 0.4049990475177765, + "learning_rate": 5.496666666666666e-05, + "loss": 0.0673, + "step": 1650 + }, + { + "epoch": 0.051237730724118774, + "grad_norm": 0.35568374395370483, + "learning_rate": 5.530000000000001e-05, + "loss": 0.0648, + "step": 1660 + }, + { + "epoch": 0.05154639175257732, + "grad_norm": 0.49035772681236267, + "learning_rate": 5.5633333333333335e-05, + "loss": 0.0723, + "step": 1670 + }, + { + "epoch": 0.05185505278103587, + "grad_norm": 0.32573720812797546, + "learning_rate": 5.596666666666667e-05, + "loss": 0.0709, + "step": 1680 + }, + { + "epoch": 0.052163713809494416, + "grad_norm": 0.5405574440956116, + "learning_rate": 5.63e-05, + "loss": 0.074, + "step": 1690 + }, + { + "epoch": 0.05247237483795296, + "grad_norm": 0.3168574571609497, + "learning_rate": 5.663333333333334e-05, + "loss": 0.0738, + "step": 1700 + }, + { + "epoch": 0.05278103586641151, + "grad_norm": 0.44639334082603455, + "learning_rate": 5.696666666666667e-05, + "loss": 0.0781, + "step": 1710 + }, + { + "epoch": 0.05308969689487005, + "grad_norm": 0.43816861510276794, + "learning_rate": 5.73e-05, + "loss": 0.0708, + "step": 1720 + }, + { + "epoch": 0.0533983579233286, + "grad_norm": 0.38822001218795776, + "learning_rate": 5.7633333333333336e-05, + "loss": 0.0685, + "step": 1730 + }, + { + "epoch": 0.053707018951787144, + "grad_norm": 0.42185550928115845, + "learning_rate": 5.796666666666667e-05, + "loss": 0.0699, + "step": 1740 + }, + { + "epoch": 0.054015679980245694, + "grad_norm": 0.5065114498138428, + "learning_rate": 5.83e-05, + "loss": 0.0768, + "step": 1750 + }, + { + "epoch": 0.054324341008704244, + "grad_norm": 0.3168993294239044, + "learning_rate": 5.863333333333334e-05, + "loss": 0.0717, + "step": 1760 + }, + { + "epoch": 0.05463300203716279, + "grad_norm": 0.357729971408844, + "learning_rate": 5.896666666666667e-05, + "loss": 0.0688, + "step": 1770 + }, + { + "epoch": 0.054941663065621336, + "grad_norm": 0.39085614681243896, + "learning_rate": 5.93e-05, + "loss": 0.069, + "step": 1780 + }, + { + "epoch": 0.05525032409407988, + "grad_norm": 0.5299559831619263, + "learning_rate": 5.9633333333333344e-05, + "loss": 0.0802, + "step": 1790 + }, + { + "epoch": 0.05555898512253843, + "grad_norm": 0.5676352977752686, + "learning_rate": 5.996666666666667e-05, + "loss": 0.0701, + "step": 1800 + }, + { + "epoch": 0.05586764615099697, + "grad_norm": 0.34953057765960693, + "learning_rate": 6.03e-05, + "loss": 0.0682, + "step": 1810 + }, + { + "epoch": 0.05617630717945552, + "grad_norm": 0.3177023231983185, + "learning_rate": 6.063333333333333e-05, + "loss": 0.0735, + "step": 1820 + }, + { + "epoch": 0.05648496820791407, + "grad_norm": 0.47239184379577637, + "learning_rate": 6.0966666666666674e-05, + "loss": 0.0673, + "step": 1830 + }, + { + "epoch": 0.056793629236372614, + "grad_norm": 0.45186981558799744, + "learning_rate": 6.13e-05, + "loss": 0.0736, + "step": 1840 + }, + { + "epoch": 0.057102290264831164, + "grad_norm": 0.2963300347328186, + "learning_rate": 6.163333333333333e-05, + "loss": 0.0765, + "step": 1850 + }, + { + "epoch": 0.05741095129328971, + "grad_norm": 0.4365266263484955, + "learning_rate": 6.196666666666668e-05, + "loss": 0.0767, + "step": 1860 + }, + { + "epoch": 0.057719612321748257, + "grad_norm": 0.37543249130249023, + "learning_rate": 6.23e-05, + "loss": 0.0675, + "step": 1870 + }, + { + "epoch": 0.058028273350206806, + "grad_norm": 0.5332486033439636, + "learning_rate": 6.263333333333333e-05, + "loss": 0.0687, + "step": 1880 + }, + { + "epoch": 0.05833693437866535, + "grad_norm": 0.4272480010986328, + "learning_rate": 6.296666666666667e-05, + "loss": 0.0762, + "step": 1890 + }, + { + "epoch": 0.0586455954071239, + "grad_norm": 0.40838587284088135, + "learning_rate": 6.330000000000001e-05, + "loss": 0.0717, + "step": 1900 + }, + { + "epoch": 0.05895425643558244, + "grad_norm": 0.48943474888801575, + "learning_rate": 6.363333333333334e-05, + "loss": 0.069, + "step": 1910 + }, + { + "epoch": 0.05926291746404099, + "grad_norm": 0.4978637099266052, + "learning_rate": 6.396666666666667e-05, + "loss": 0.0736, + "step": 1920 + }, + { + "epoch": 0.059571578492499534, + "grad_norm": 0.38876011967658997, + "learning_rate": 6.43e-05, + "loss": 0.0715, + "step": 1930 + }, + { + "epoch": 0.059880239520958084, + "grad_norm": 0.3277791142463684, + "learning_rate": 6.463333333333334e-05, + "loss": 0.0715, + "step": 1940 + }, + { + "epoch": 0.060188900549416634, + "grad_norm": 0.3198246955871582, + "learning_rate": 6.496666666666667e-05, + "loss": 0.0699, + "step": 1950 + }, + { + "epoch": 0.06049756157787518, + "grad_norm": 0.4274479150772095, + "learning_rate": 6.53e-05, + "loss": 0.0725, + "step": 1960 + }, + { + "epoch": 0.060806222606333726, + "grad_norm": 0.4581236243247986, + "learning_rate": 6.563333333333333e-05, + "loss": 0.0727, + "step": 1970 + }, + { + "epoch": 0.06111488363479227, + "grad_norm": 0.4974837899208069, + "learning_rate": 6.596666666666667e-05, + "loss": 0.0716, + "step": 1980 + }, + { + "epoch": 0.06142354466325082, + "grad_norm": 0.37381511926651, + "learning_rate": 6.630000000000001e-05, + "loss": 0.0732, + "step": 1990 + }, + { + "epoch": 0.06173220569170936, + "grad_norm": 0.40079811215400696, + "learning_rate": 6.663333333333333e-05, + "loss": 0.0713, + "step": 2000 + }, + { + "epoch": 0.06204086672016791, + "grad_norm": 0.5098493099212646, + "learning_rate": 6.696666666666666e-05, + "loss": 0.0709, + "step": 2010 + }, + { + "epoch": 0.06234952774862646, + "grad_norm": 0.6400183439254761, + "learning_rate": 6.730000000000001e-05, + "loss": 0.0796, + "step": 2020 + }, + { + "epoch": 0.06265818877708501, + "grad_norm": 0.329706072807312, + "learning_rate": 6.763333333333334e-05, + "loss": 0.0723, + "step": 2030 + }, + { + "epoch": 0.06296684980554355, + "grad_norm": 0.3734639883041382, + "learning_rate": 6.796666666666666e-05, + "loss": 0.0671, + "step": 2040 + }, + { + "epoch": 0.0632755108340021, + "grad_norm": 0.401265949010849, + "learning_rate": 6.83e-05, + "loss": 0.0752, + "step": 2050 + }, + { + "epoch": 0.06358417186246064, + "grad_norm": 0.46421724557876587, + "learning_rate": 6.863333333333334e-05, + "loss": 0.0726, + "step": 2060 + }, + { + "epoch": 0.0638928328909192, + "grad_norm": 0.5074324011802673, + "learning_rate": 6.896666666666667e-05, + "loss": 0.0758, + "step": 2070 + }, + { + "epoch": 0.06420149391937774, + "grad_norm": 0.36859655380249023, + "learning_rate": 6.93e-05, + "loss": 0.0648, + "step": 2080 + }, + { + "epoch": 0.06451015494783628, + "grad_norm": 0.3827269971370697, + "learning_rate": 6.963333333333334e-05, + "loss": 0.0751, + "step": 2090 + }, + { + "epoch": 0.06481881597629484, + "grad_norm": 0.4487954378128052, + "learning_rate": 6.996666666666667e-05, + "loss": 0.0713, + "step": 2100 + }, + { + "epoch": 0.06512747700475338, + "grad_norm": 0.4430448114871979, + "learning_rate": 7.03e-05, + "loss": 0.0649, + "step": 2110 + }, + { + "epoch": 0.06543613803321192, + "grad_norm": 0.29798319935798645, + "learning_rate": 7.063333333333333e-05, + "loss": 0.0684, + "step": 2120 + }, + { + "epoch": 0.06574479906167047, + "grad_norm": 0.5166922807693481, + "learning_rate": 7.096666666666667e-05, + "loss": 0.0754, + "step": 2130 + }, + { + "epoch": 0.06605346009012902, + "grad_norm": 0.407734215259552, + "learning_rate": 7.13e-05, + "loss": 0.0684, + "step": 2140 + }, + { + "epoch": 0.06636212111858757, + "grad_norm": 0.4556084871292114, + "learning_rate": 7.163333333333334e-05, + "loss": 0.0712, + "step": 2150 + }, + { + "epoch": 0.06667078214704611, + "grad_norm": 0.5520690679550171, + "learning_rate": 7.196666666666668e-05, + "loss": 0.0693, + "step": 2160 + }, + { + "epoch": 0.06697944317550467, + "grad_norm": 0.5128118991851807, + "learning_rate": 7.23e-05, + "loss": 0.069, + "step": 2170 + }, + { + "epoch": 0.06728810420396321, + "grad_norm": 0.47830867767333984, + "learning_rate": 7.263333333333334e-05, + "loss": 0.0698, + "step": 2180 + }, + { + "epoch": 0.06759676523242175, + "grad_norm": 0.4243362247943878, + "learning_rate": 7.296666666666667e-05, + "loss": 0.0683, + "step": 2190 + }, + { + "epoch": 0.0679054262608803, + "grad_norm": 0.5219701528549194, + "learning_rate": 7.33e-05, + "loss": 0.0749, + "step": 2200 + }, + { + "epoch": 0.06821408728933885, + "grad_norm": 0.30258312821388245, + "learning_rate": 7.363333333333334e-05, + "loss": 0.0753, + "step": 2210 + }, + { + "epoch": 0.0685227483177974, + "grad_norm": 0.410013347864151, + "learning_rate": 7.396666666666667e-05, + "loss": 0.0741, + "step": 2220 + }, + { + "epoch": 0.06883140934625594, + "grad_norm": 0.3756735622882843, + "learning_rate": 7.43e-05, + "loss": 0.068, + "step": 2230 + }, + { + "epoch": 0.0691400703747145, + "grad_norm": 0.43535253405570984, + "learning_rate": 7.463333333333334e-05, + "loss": 0.0703, + "step": 2240 + }, + { + "epoch": 0.06944873140317304, + "grad_norm": 0.5140138864517212, + "learning_rate": 7.496666666666667e-05, + "loss": 0.0719, + "step": 2250 + }, + { + "epoch": 0.06975739243163158, + "grad_norm": 0.43354108929634094, + "learning_rate": 7.53e-05, + "loss": 0.0749, + "step": 2260 + }, + { + "epoch": 0.07006605346009014, + "grad_norm": 0.35576584935188293, + "learning_rate": 7.563333333333333e-05, + "loss": 0.0708, + "step": 2270 + }, + { + "epoch": 0.07037471448854868, + "grad_norm": 0.3253732919692993, + "learning_rate": 7.596666666666668e-05, + "loss": 0.0638, + "step": 2280 + }, + { + "epoch": 0.07068337551700722, + "grad_norm": 0.3316190540790558, + "learning_rate": 7.630000000000001e-05, + "loss": 0.0718, + "step": 2290 + }, + { + "epoch": 0.07099203654546576, + "grad_norm": 0.3668586015701294, + "learning_rate": 7.663333333333333e-05, + "loss": 0.0701, + "step": 2300 + }, + { + "epoch": 0.07130069757392432, + "grad_norm": 0.3885926902294159, + "learning_rate": 7.696666666666668e-05, + "loss": 0.0726, + "step": 2310 + }, + { + "epoch": 0.07160935860238286, + "grad_norm": 0.30676788091659546, + "learning_rate": 7.730000000000001e-05, + "loss": 0.0753, + "step": 2320 + }, + { + "epoch": 0.0719180196308414, + "grad_norm": 0.38624000549316406, + "learning_rate": 7.763333333333334e-05, + "loss": 0.0736, + "step": 2330 + }, + { + "epoch": 0.07222668065929996, + "grad_norm": 0.4739297330379486, + "learning_rate": 7.796666666666666e-05, + "loss": 0.0666, + "step": 2340 + }, + { + "epoch": 0.0725353416877585, + "grad_norm": 0.4425777196884155, + "learning_rate": 7.83e-05, + "loss": 0.0689, + "step": 2350 + }, + { + "epoch": 0.07284400271621705, + "grad_norm": 0.34028005599975586, + "learning_rate": 7.863333333333334e-05, + "loss": 0.0742, + "step": 2360 + }, + { + "epoch": 0.07315266374467559, + "grad_norm": 0.41411375999450684, + "learning_rate": 7.896666666666667e-05, + "loss": 0.068, + "step": 2370 + }, + { + "epoch": 0.07346132477313415, + "grad_norm": 0.3086642920970917, + "learning_rate": 7.93e-05, + "loss": 0.0684, + "step": 2380 + }, + { + "epoch": 0.07376998580159269, + "grad_norm": 0.34911349415779114, + "learning_rate": 7.963333333333334e-05, + "loss": 0.069, + "step": 2390 + }, + { + "epoch": 0.07407864683005123, + "grad_norm": 0.30732566118240356, + "learning_rate": 7.996666666666667e-05, + "loss": 0.0644, + "step": 2400 + }, + { + "epoch": 0.07438730785850979, + "grad_norm": 0.4057803750038147, + "learning_rate": 8.030000000000001e-05, + "loss": 0.0657, + "step": 2410 + }, + { + "epoch": 0.07469596888696833, + "grad_norm": 0.35782232880592346, + "learning_rate": 8.063333333333333e-05, + "loss": 0.076, + "step": 2420 + }, + { + "epoch": 0.07500462991542688, + "grad_norm": 0.3496623635292053, + "learning_rate": 8.096666666666667e-05, + "loss": 0.0689, + "step": 2430 + }, + { + "epoch": 0.07531329094388542, + "grad_norm": 0.3068278133869171, + "learning_rate": 8.13e-05, + "loss": 0.0679, + "step": 2440 + }, + { + "epoch": 0.07562195197234398, + "grad_norm": 0.6721469759941101, + "learning_rate": 8.163333333333334e-05, + "loss": 0.0667, + "step": 2450 + }, + { + "epoch": 0.07593061300080252, + "grad_norm": 0.3017984628677368, + "learning_rate": 8.196666666666668e-05, + "loss": 0.07, + "step": 2460 + }, + { + "epoch": 0.07623927402926106, + "grad_norm": 0.3871825933456421, + "learning_rate": 8.23e-05, + "loss": 0.0703, + "step": 2470 + }, + { + "epoch": 0.07654793505771962, + "grad_norm": 0.34531816840171814, + "learning_rate": 8.263333333333334e-05, + "loss": 0.0644, + "step": 2480 + }, + { + "epoch": 0.07685659608617816, + "grad_norm": 0.2886400520801544, + "learning_rate": 8.296666666666667e-05, + "loss": 0.0704, + "step": 2490 + }, + { + "epoch": 0.0771652571146367, + "grad_norm": 0.37211427092552185, + "learning_rate": 8.33e-05, + "loss": 0.0716, + "step": 2500 + }, + { + "epoch": 0.07747391814309525, + "grad_norm": 0.27726224064826965, + "learning_rate": 8.363333333333334e-05, + "loss": 0.0687, + "step": 2510 + }, + { + "epoch": 0.0777825791715538, + "grad_norm": 1.1890395879745483, + "learning_rate": 8.396666666666667e-05, + "loss": 0.0687, + "step": 2520 + }, + { + "epoch": 0.07809124020001235, + "grad_norm": 0.48000407218933105, + "learning_rate": 8.43e-05, + "loss": 0.0739, + "step": 2530 + }, + { + "epoch": 0.07839990122847089, + "grad_norm": 0.4494258463382721, + "learning_rate": 8.463333333333335e-05, + "loss": 0.0676, + "step": 2540 + }, + { + "epoch": 0.07870856225692945, + "grad_norm": 0.39537760615348816, + "learning_rate": 8.496666666666667e-05, + "loss": 0.0757, + "step": 2550 + }, + { + "epoch": 0.07901722328538799, + "grad_norm": 0.4127436578273773, + "learning_rate": 8.53e-05, + "loss": 0.0746, + "step": 2560 + }, + { + "epoch": 0.07932588431384653, + "grad_norm": 0.3919200897216797, + "learning_rate": 8.563333333333333e-05, + "loss": 0.0697, + "step": 2570 + }, + { + "epoch": 0.07963454534230507, + "grad_norm": 0.3531115651130676, + "learning_rate": 8.596666666666668e-05, + "loss": 0.068, + "step": 2580 + }, + { + "epoch": 0.07994320637076363, + "grad_norm": 0.4130890667438507, + "learning_rate": 8.63e-05, + "loss": 0.0636, + "step": 2590 + }, + { + "epoch": 0.08025186739922217, + "grad_norm": 0.3101487159729004, + "learning_rate": 8.663333333333333e-05, + "loss": 0.0711, + "step": 2600 + }, + { + "epoch": 0.08056052842768072, + "grad_norm": 0.5342437624931335, + "learning_rate": 8.696666666666668e-05, + "loss": 0.0706, + "step": 2610 + }, + { + "epoch": 0.08086918945613927, + "grad_norm": 0.44725316762924194, + "learning_rate": 8.730000000000001e-05, + "loss": 0.0735, + "step": 2620 + }, + { + "epoch": 0.08117785048459782, + "grad_norm": 0.2709047198295593, + "learning_rate": 8.763333333333334e-05, + "loss": 0.0702, + "step": 2630 + }, + { + "epoch": 0.08148651151305636, + "grad_norm": 0.37155625224113464, + "learning_rate": 8.796666666666667e-05, + "loss": 0.0696, + "step": 2640 + }, + { + "epoch": 0.0817951725415149, + "grad_norm": 0.46019795536994934, + "learning_rate": 8.83e-05, + "loss": 0.0724, + "step": 2650 + }, + { + "epoch": 0.08210383356997346, + "grad_norm": 0.3714541494846344, + "learning_rate": 8.863333333333334e-05, + "loss": 0.0715, + "step": 2660 + }, + { + "epoch": 0.082412494598432, + "grad_norm": 0.3796297013759613, + "learning_rate": 8.896666666666667e-05, + "loss": 0.0689, + "step": 2670 + }, + { + "epoch": 0.08272115562689054, + "grad_norm": 0.33451229333877563, + "learning_rate": 8.93e-05, + "loss": 0.0682, + "step": 2680 + }, + { + "epoch": 0.0830298166553491, + "grad_norm": 0.37449976801872253, + "learning_rate": 8.963333333333333e-05, + "loss": 0.0687, + "step": 2690 + }, + { + "epoch": 0.08333847768380764, + "grad_norm": 0.43841490149497986, + "learning_rate": 8.996666666666667e-05, + "loss": 0.0705, + "step": 2700 + }, + { + "epoch": 0.08364713871226619, + "grad_norm": 0.47989708185195923, + "learning_rate": 9.030000000000001e-05, + "loss": 0.072, + "step": 2710 + }, + { + "epoch": 0.08395579974072473, + "grad_norm": 0.4006040096282959, + "learning_rate": 9.063333333333333e-05, + "loss": 0.0703, + "step": 2720 + }, + { + "epoch": 0.08426446076918329, + "grad_norm": 0.3508071005344391, + "learning_rate": 9.096666666666666e-05, + "loss": 0.0718, + "step": 2730 + }, + { + "epoch": 0.08457312179764183, + "grad_norm": 0.3814544677734375, + "learning_rate": 9.130000000000001e-05, + "loss": 0.0675, + "step": 2740 + }, + { + "epoch": 0.08488178282610037, + "grad_norm": 0.43475764989852905, + "learning_rate": 9.163333333333334e-05, + "loss": 0.0623, + "step": 2750 + }, + { + "epoch": 0.08519044385455893, + "grad_norm": 0.32663336396217346, + "learning_rate": 9.196666666666666e-05, + "loss": 0.0623, + "step": 2760 + }, + { + "epoch": 0.08549910488301747, + "grad_norm": 0.37711772322654724, + "learning_rate": 9.230000000000001e-05, + "loss": 0.0671, + "step": 2770 + }, + { + "epoch": 0.08580776591147601, + "grad_norm": 0.35686421394348145, + "learning_rate": 9.263333333333334e-05, + "loss": 0.0689, + "step": 2780 + }, + { + "epoch": 0.08611642693993457, + "grad_norm": 0.32114270329475403, + "learning_rate": 9.296666666666667e-05, + "loss": 0.0709, + "step": 2790 + }, + { + "epoch": 0.08642508796839311, + "grad_norm": 0.3517875075340271, + "learning_rate": 9.33e-05, + "loss": 0.0714, + "step": 2800 + }, + { + "epoch": 0.08673374899685166, + "grad_norm": 0.3057754337787628, + "learning_rate": 9.363333333333334e-05, + "loss": 0.0691, + "step": 2810 + }, + { + "epoch": 0.0870424100253102, + "grad_norm": 0.4234413504600525, + "learning_rate": 9.396666666666667e-05, + "loss": 0.0658, + "step": 2820 + }, + { + "epoch": 0.08735107105376876, + "grad_norm": 0.3610641658306122, + "learning_rate": 9.43e-05, + "loss": 0.0663, + "step": 2830 + }, + { + "epoch": 0.0876597320822273, + "grad_norm": 0.38690486550331116, + "learning_rate": 9.463333333333333e-05, + "loss": 0.0679, + "step": 2840 + }, + { + "epoch": 0.08796839311068584, + "grad_norm": 0.43143901228904724, + "learning_rate": 9.496666666666667e-05, + "loss": 0.0704, + "step": 2850 + }, + { + "epoch": 0.0882770541391444, + "grad_norm": 0.3732384741306305, + "learning_rate": 9.53e-05, + "loss": 0.0764, + "step": 2860 + }, + { + "epoch": 0.08858571516760294, + "grad_norm": 0.20282624661922455, + "learning_rate": 9.563333333333334e-05, + "loss": 0.0685, + "step": 2870 + }, + { + "epoch": 0.08889437619606148, + "grad_norm": 0.518351674079895, + "learning_rate": 9.596666666666668e-05, + "loss": 0.0702, + "step": 2880 + }, + { + "epoch": 0.08920303722452003, + "grad_norm": 0.4288371801376343, + "learning_rate": 9.63e-05, + "loss": 0.072, + "step": 2890 + }, + { + "epoch": 0.08951169825297858, + "grad_norm": 0.299905002117157, + "learning_rate": 9.663333333333334e-05, + "loss": 0.0652, + "step": 2900 + }, + { + "epoch": 0.08982035928143713, + "grad_norm": 0.3006996512413025, + "learning_rate": 9.696666666666667e-05, + "loss": 0.0668, + "step": 2910 + }, + { + "epoch": 0.09012902030989567, + "grad_norm": 0.37174516916275024, + "learning_rate": 9.730000000000001e-05, + "loss": 0.0742, + "step": 2920 + }, + { + "epoch": 0.09043768133835423, + "grad_norm": 1.0711703300476074, + "learning_rate": 9.763333333333334e-05, + "loss": 0.0705, + "step": 2930 + }, + { + "epoch": 0.09074634236681277, + "grad_norm": 0.5607337951660156, + "learning_rate": 9.796666666666667e-05, + "loss": 0.0755, + "step": 2940 + }, + { + "epoch": 0.09105500339527131, + "grad_norm": 0.38709041476249695, + "learning_rate": 9.83e-05, + "loss": 0.0711, + "step": 2950 + }, + { + "epoch": 0.09136366442372985, + "grad_norm": 0.39903029799461365, + "learning_rate": 9.863333333333334e-05, + "loss": 0.0687, + "step": 2960 + }, + { + "epoch": 0.09167232545218841, + "grad_norm": 0.374950110912323, + "learning_rate": 9.896666666666667e-05, + "loss": 0.0664, + "step": 2970 + }, + { + "epoch": 0.09198098648064695, + "grad_norm": 0.39626064896583557, + "learning_rate": 9.93e-05, + "loss": 0.0703, + "step": 2980 + }, + { + "epoch": 0.0922896475091055, + "grad_norm": 0.42498621344566345, + "learning_rate": 9.963333333333333e-05, + "loss": 0.0672, + "step": 2990 + }, + { + "epoch": 0.09259830853756405, + "grad_norm": 0.436946839094162, + "learning_rate": 9.996666666666668e-05, + "loss": 0.0685, + "step": 3000 + }, + { + "epoch": 0.0929069695660226, + "grad_norm": 0.3673728108406067, + "learning_rate": 9.999999384858465e-05, + "loss": 0.0658, + "step": 3010 + }, + { + "epoch": 0.09321563059448114, + "grad_norm": 0.30466368794441223, + "learning_rate": 9.999997258443473e-05, + "loss": 0.0691, + "step": 3020 + }, + { + "epoch": 0.09352429162293968, + "grad_norm": 0.3451826572418213, + "learning_rate": 9.999993613161331e-05, + "loss": 0.066, + "step": 3030 + }, + { + "epoch": 0.09383295265139824, + "grad_norm": 0.2385428249835968, + "learning_rate": 9.999988449013146e-05, + "loss": 0.069, + "step": 3040 + }, + { + "epoch": 0.09414161367985678, + "grad_norm": 0.36629945039749146, + "learning_rate": 9.99998176600049e-05, + "loss": 0.0655, + "step": 3050 + }, + { + "epoch": 0.09445027470831532, + "grad_norm": 0.27359941601753235, + "learning_rate": 9.999973564125389e-05, + "loss": 0.0659, + "step": 3060 + }, + { + "epoch": 0.09475893573677388, + "grad_norm": 0.3932338356971741, + "learning_rate": 9.999963843390335e-05, + "loss": 0.0667, + "step": 3070 + }, + { + "epoch": 0.09506759676523242, + "grad_norm": 0.32883819937705994, + "learning_rate": 9.999952603798282e-05, + "loss": 0.0649, + "step": 3080 + }, + { + "epoch": 0.09537625779369097, + "grad_norm": 0.4035113453865051, + "learning_rate": 9.999939845352646e-05, + "loss": 0.0718, + "step": 3090 + }, + { + "epoch": 0.09568491882214951, + "grad_norm": 0.2801380753517151, + "learning_rate": 9.999925568057298e-05, + "loss": 0.0738, + "step": 3100 + }, + { + "epoch": 0.09599357985060807, + "grad_norm": 0.3102757930755615, + "learning_rate": 9.999909771916578e-05, + "loss": 0.0674, + "step": 3110 + }, + { + "epoch": 0.09630224087906661, + "grad_norm": 0.4002803862094879, + "learning_rate": 9.999892456935285e-05, + "loss": 0.0712, + "step": 3120 + }, + { + "epoch": 0.09661090190752515, + "grad_norm": 0.9150711297988892, + "learning_rate": 9.999873623118679e-05, + "loss": 0.071, + "step": 3130 + }, + { + "epoch": 0.09691956293598371, + "grad_norm": 0.5354248881340027, + "learning_rate": 9.999853270472479e-05, + "loss": 0.072, + "step": 3140 + }, + { + "epoch": 0.09722822396444225, + "grad_norm": 0.3699294924736023, + "learning_rate": 9.999831399002871e-05, + "loss": 0.0673, + "step": 3150 + }, + { + "epoch": 0.0975368849929008, + "grad_norm": 0.34525948762893677, + "learning_rate": 9.999808008716494e-05, + "loss": 0.0687, + "step": 3160 + }, + { + "epoch": 0.09784554602135934, + "grad_norm": 0.366803914308548, + "learning_rate": 9.999783099620459e-05, + "loss": 0.0702, + "step": 3170 + }, + { + "epoch": 0.0981542070498179, + "grad_norm": 0.3192853033542633, + "learning_rate": 9.999756671722328e-05, + "loss": 0.0657, + "step": 3180 + }, + { + "epoch": 0.09846286807827644, + "grad_norm": 0.30014878511428833, + "learning_rate": 9.99972872503013e-05, + "loss": 0.0675, + "step": 3190 + }, + { + "epoch": 0.09877152910673498, + "grad_norm": 0.6571142673492432, + "learning_rate": 9.999699259552359e-05, + "loss": 0.0697, + "step": 3200 + }, + { + "epoch": 0.09908019013519354, + "grad_norm": 0.44107186794281006, + "learning_rate": 9.99966827529796e-05, + "loss": 0.071, + "step": 3210 + }, + { + "epoch": 0.09938885116365208, + "grad_norm": 0.27173668146133423, + "learning_rate": 9.999635772276348e-05, + "loss": 0.0662, + "step": 3220 + }, + { + "epoch": 0.09969751219211062, + "grad_norm": 0.3273988664150238, + "learning_rate": 9.999601750497396e-05, + "loss": 0.0667, + "step": 3230 + }, + { + "epoch": 0.10000617322056918, + "grad_norm": 0.2922723889350891, + "learning_rate": 9.99956620997144e-05, + "loss": 0.0704, + "step": 3240 + }, + { + "epoch": 0.10031483424902772, + "grad_norm": 0.3041306138038635, + "learning_rate": 9.999529150709275e-05, + "loss": 0.0686, + "step": 3250 + }, + { + "epoch": 0.10062349527748626, + "grad_norm": 0.3177841603755951, + "learning_rate": 9.999490572722158e-05, + "loss": 0.068, + "step": 3260 + }, + { + "epoch": 0.1009321563059448, + "grad_norm": 0.27343833446502686, + "learning_rate": 9.99945047602181e-05, + "loss": 0.0716, + "step": 3270 + }, + { + "epoch": 0.10124081733440336, + "grad_norm": 0.39326339960098267, + "learning_rate": 9.99940886062041e-05, + "loss": 0.0685, + "step": 3280 + }, + { + "epoch": 0.1015494783628619, + "grad_norm": 0.2710900902748108, + "learning_rate": 9.999365726530599e-05, + "loss": 0.0774, + "step": 3290 + }, + { + "epoch": 0.10185813939132045, + "grad_norm": 0.3369479775428772, + "learning_rate": 9.999321073765481e-05, + "loss": 0.0682, + "step": 3300 + }, + { + "epoch": 0.102166800419779, + "grad_norm": 0.293112576007843, + "learning_rate": 9.99927490233862e-05, + "loss": 0.0688, + "step": 3310 + }, + { + "epoch": 0.10247546144823755, + "grad_norm": 0.4611547589302063, + "learning_rate": 9.999227212264043e-05, + "loss": 0.0705, + "step": 3320 + }, + { + "epoch": 0.10278412247669609, + "grad_norm": 0.2801327109336853, + "learning_rate": 9.999178003556236e-05, + "loss": 0.064, + "step": 3330 + }, + { + "epoch": 0.10309278350515463, + "grad_norm": 0.4939819276332855, + "learning_rate": 9.999127276230146e-05, + "loss": 0.0702, + "step": 3340 + }, + { + "epoch": 0.10340144453361319, + "grad_norm": 0.27057623863220215, + "learning_rate": 9.999075030301184e-05, + "loss": 0.0651, + "step": 3350 + }, + { + "epoch": 0.10371010556207173, + "grad_norm": 0.31387823820114136, + "learning_rate": 9.999021265785221e-05, + "loss": 0.0674, + "step": 3360 + }, + { + "epoch": 0.10401876659053028, + "grad_norm": 0.4366486370563507, + "learning_rate": 9.998965982698589e-05, + "loss": 0.073, + "step": 3370 + }, + { + "epoch": 0.10432742761898883, + "grad_norm": 0.41133832931518555, + "learning_rate": 9.998909181058082e-05, + "loss": 0.0657, + "step": 3380 + }, + { + "epoch": 0.10463608864744738, + "grad_norm": 0.7902501225471497, + "learning_rate": 9.998850860880953e-05, + "loss": 0.07, + "step": 3390 + }, + { + "epoch": 0.10494474967590592, + "grad_norm": 0.34875214099884033, + "learning_rate": 9.998791022184922e-05, + "loss": 0.0696, + "step": 3400 + }, + { + "epoch": 0.10525341070436446, + "grad_norm": 0.3100384771823883, + "learning_rate": 9.99872966498816e-05, + "loss": 0.0632, + "step": 3410 + }, + { + "epoch": 0.10556207173282302, + "grad_norm": 0.4046914875507355, + "learning_rate": 9.998666789309313e-05, + "loss": 0.0656, + "step": 3420 + }, + { + "epoch": 0.10587073276128156, + "grad_norm": 0.30005940794944763, + "learning_rate": 9.998602395167475e-05, + "loss": 0.0628, + "step": 3430 + }, + { + "epoch": 0.1061793937897401, + "grad_norm": 0.3829677104949951, + "learning_rate": 9.998536482582213e-05, + "loss": 0.0715, + "step": 3440 + }, + { + "epoch": 0.10648805481819866, + "grad_norm": 0.37534213066101074, + "learning_rate": 9.998469051573544e-05, + "loss": 0.0675, + "step": 3450 + }, + { + "epoch": 0.1067967158466572, + "grad_norm": 1.118354320526123, + "learning_rate": 9.998400102161954e-05, + "loss": 0.071, + "step": 3460 + }, + { + "epoch": 0.10710537687511575, + "grad_norm": 0.2821442484855652, + "learning_rate": 9.998329634368388e-05, + "loss": 0.0615, + "step": 3470 + }, + { + "epoch": 0.10741403790357429, + "grad_norm": 0.4025896489620209, + "learning_rate": 9.998257648214253e-05, + "loss": 0.0703, + "step": 3480 + }, + { + "epoch": 0.10772269893203285, + "grad_norm": 0.30867743492126465, + "learning_rate": 9.998184143721417e-05, + "loss": 0.0643, + "step": 3490 + }, + { + "epoch": 0.10803135996049139, + "grad_norm": 0.43436482548713684, + "learning_rate": 9.998109120912206e-05, + "loss": 0.0673, + "step": 3500 + }, + { + "epoch": 0.10834002098894993, + "grad_norm": 0.37062472105026245, + "learning_rate": 9.998032579809411e-05, + "loss": 0.0645, + "step": 3510 + }, + { + "epoch": 0.10864868201740849, + "grad_norm": 0.3269613981246948, + "learning_rate": 9.997954520436286e-05, + "loss": 0.0698, + "step": 3520 + }, + { + "epoch": 0.10895734304586703, + "grad_norm": 0.5092619061470032, + "learning_rate": 9.997874942816538e-05, + "loss": 0.069, + "step": 3530 + }, + { + "epoch": 0.10926600407432557, + "grad_norm": 0.42629948258399963, + "learning_rate": 9.997793846974345e-05, + "loss": 0.0671, + "step": 3540 + }, + { + "epoch": 0.10957466510278412, + "grad_norm": 0.3027883470058441, + "learning_rate": 9.997711232934341e-05, + "loss": 0.0675, + "step": 3550 + }, + { + "epoch": 0.10988332613124267, + "grad_norm": 0.2914540469646454, + "learning_rate": 9.99762710072162e-05, + "loss": 0.0675, + "step": 3560 + }, + { + "epoch": 0.11019198715970122, + "grad_norm": 0.30372536182403564, + "learning_rate": 9.997541450361743e-05, + "loss": 0.0653, + "step": 3570 + }, + { + "epoch": 0.11050064818815976, + "grad_norm": 0.2790224850177765, + "learning_rate": 9.997454281880723e-05, + "loss": 0.0704, + "step": 3580 + }, + { + "epoch": 0.11080930921661832, + "grad_norm": 0.3698699474334717, + "learning_rate": 9.997365595305044e-05, + "loss": 0.0689, + "step": 3590 + }, + { + "epoch": 0.11111797024507686, + "grad_norm": 0.34927141666412354, + "learning_rate": 9.997275390661644e-05, + "loss": 0.0708, + "step": 3600 + }, + { + "epoch": 0.1114266312735354, + "grad_norm": 0.3716798424720764, + "learning_rate": 9.997183667977926e-05, + "loss": 0.0665, + "step": 3610 + }, + { + "epoch": 0.11173529230199394, + "grad_norm": 0.39003702998161316, + "learning_rate": 9.997090427281752e-05, + "loss": 0.0697, + "step": 3620 + }, + { + "epoch": 0.1120439533304525, + "grad_norm": 0.3676378130912781, + "learning_rate": 9.996995668601448e-05, + "loss": 0.0628, + "step": 3630 + }, + { + "epoch": 0.11235261435891104, + "grad_norm": 0.49715468287467957, + "learning_rate": 9.996899391965798e-05, + "loss": 0.0684, + "step": 3640 + }, + { + "epoch": 0.11266127538736959, + "grad_norm": 0.288303017616272, + "learning_rate": 9.996801597404048e-05, + "loss": 0.0672, + "step": 3650 + }, + { + "epoch": 0.11296993641582814, + "grad_norm": 0.38992705941200256, + "learning_rate": 9.996702284945905e-05, + "loss": 0.0649, + "step": 3660 + }, + { + "epoch": 0.11327859744428669, + "grad_norm": 0.3869478702545166, + "learning_rate": 9.996601454621539e-05, + "loss": 0.0648, + "step": 3670 + }, + { + "epoch": 0.11358725847274523, + "grad_norm": 0.4517019987106323, + "learning_rate": 9.996499106461577e-05, + "loss": 0.0633, + "step": 3680 + }, + { + "epoch": 0.11389591950120377, + "grad_norm": 0.31691229343414307, + "learning_rate": 9.996395240497112e-05, + "loss": 0.0702, + "step": 3690 + }, + { + "epoch": 0.11420458052966233, + "grad_norm": 0.3217264413833618, + "learning_rate": 9.996289856759696e-05, + "loss": 0.0679, + "step": 3700 + }, + { + "epoch": 0.11451324155812087, + "grad_norm": 0.3769218921661377, + "learning_rate": 9.996182955281342e-05, + "loss": 0.068, + "step": 3710 + }, + { + "epoch": 0.11482190258657941, + "grad_norm": 0.3845515847206116, + "learning_rate": 9.996074536094519e-05, + "loss": 0.0649, + "step": 3720 + }, + { + "epoch": 0.11513056361503797, + "grad_norm": 0.43319809436798096, + "learning_rate": 9.995964599232168e-05, + "loss": 0.0645, + "step": 3730 + }, + { + "epoch": 0.11543922464349651, + "grad_norm": 0.26069578528404236, + "learning_rate": 9.995853144727683e-05, + "loss": 0.069, + "step": 3740 + }, + { + "epoch": 0.11574788567195506, + "grad_norm": 0.2768746018409729, + "learning_rate": 9.99574017261492e-05, + "loss": 0.0705, + "step": 3750 + }, + { + "epoch": 0.11605654670041361, + "grad_norm": 0.3470922112464905, + "learning_rate": 9.995625682928198e-05, + "loss": 0.064, + "step": 3760 + }, + { + "epoch": 0.11636520772887216, + "grad_norm": 0.3051047623157501, + "learning_rate": 9.995509675702295e-05, + "loss": 0.0665, + "step": 3770 + }, + { + "epoch": 0.1166738687573307, + "grad_norm": 0.3940667510032654, + "learning_rate": 9.995392150972451e-05, + "loss": 0.0643, + "step": 3780 + }, + { + "epoch": 0.11698252978578924, + "grad_norm": 0.3864079713821411, + "learning_rate": 9.995273108774366e-05, + "loss": 0.0677, + "step": 3790 + }, + { + "epoch": 0.1172911908142478, + "grad_norm": 0.2979573905467987, + "learning_rate": 9.995152549144205e-05, + "loss": 0.0675, + "step": 3800 + }, + { + "epoch": 0.11759985184270634, + "grad_norm": 0.3408135175704956, + "learning_rate": 9.995030472118587e-05, + "loss": 0.0647, + "step": 3810 + }, + { + "epoch": 0.11790851287116488, + "grad_norm": 0.32723891735076904, + "learning_rate": 9.9949068777346e-05, + "loss": 0.067, + "step": 3820 + }, + { + "epoch": 0.11821717389962344, + "grad_norm": 0.7935934662818909, + "learning_rate": 9.994781766029786e-05, + "loss": 0.0668, + "step": 3830 + }, + { + "epoch": 0.11852583492808198, + "grad_norm": 0.3368012011051178, + "learning_rate": 9.994655137042151e-05, + "loss": 0.0637, + "step": 3840 + }, + { + "epoch": 0.11883449595654053, + "grad_norm": 0.34537455439567566, + "learning_rate": 9.99452699081016e-05, + "loss": 0.0642, + "step": 3850 + }, + { + "epoch": 0.11914315698499907, + "grad_norm": 0.2920161485671997, + "learning_rate": 9.994397327372743e-05, + "loss": 0.0689, + "step": 3860 + }, + { + "epoch": 0.11945181801345763, + "grad_norm": 0.3744475245475769, + "learning_rate": 9.994266146769286e-05, + "loss": 0.0666, + "step": 3870 + }, + { + "epoch": 0.11976047904191617, + "grad_norm": 0.33258703351020813, + "learning_rate": 9.994133449039642e-05, + "loss": 0.0622, + "step": 3880 + }, + { + "epoch": 0.12006914007037471, + "grad_norm": 0.3355514705181122, + "learning_rate": 9.993999234224118e-05, + "loss": 0.0604, + "step": 3890 + }, + { + "epoch": 0.12037780109883327, + "grad_norm": 0.3292757570743561, + "learning_rate": 9.993863502363485e-05, + "loss": 0.0611, + "step": 3900 + }, + { + "epoch": 0.12068646212729181, + "grad_norm": 0.4871256649494171, + "learning_rate": 9.993726253498976e-05, + "loss": 0.0677, + "step": 3910 + }, + { + "epoch": 0.12099512315575035, + "grad_norm": 0.34734606742858887, + "learning_rate": 9.993587487672282e-05, + "loss": 0.0688, + "step": 3920 + }, + { + "epoch": 0.1213037841842089, + "grad_norm": 0.3722241520881653, + "learning_rate": 9.993447204925558e-05, + "loss": 0.0622, + "step": 3930 + }, + { + "epoch": 0.12161244521266745, + "grad_norm": 0.28459739685058594, + "learning_rate": 9.993305405301416e-05, + "loss": 0.0635, + "step": 3940 + }, + { + "epoch": 0.121921106241126, + "grad_norm": 0.2941494584083557, + "learning_rate": 9.993162088842935e-05, + "loss": 0.0641, + "step": 3950 + }, + { + "epoch": 0.12222976726958454, + "grad_norm": 0.2894752025604248, + "learning_rate": 9.993017255593646e-05, + "loss": 0.0702, + "step": 3960 + }, + { + "epoch": 0.1225384282980431, + "grad_norm": 0.29609984159469604, + "learning_rate": 9.992870905597548e-05, + "loss": 0.0664, + "step": 3970 + }, + { + "epoch": 0.12284708932650164, + "grad_norm": 0.361286997795105, + "learning_rate": 9.9927230388991e-05, + "loss": 0.0735, + "step": 3980 + }, + { + "epoch": 0.12315575035496018, + "grad_norm": 0.4635210335254669, + "learning_rate": 9.992573655543215e-05, + "loss": 0.0732, + "step": 3990 + }, + { + "epoch": 0.12346441138341872, + "grad_norm": 0.3207724094390869, + "learning_rate": 9.992422755575277e-05, + "loss": 0.0645, + "step": 4000 + }, + { + "epoch": 0.12377307241187728, + "grad_norm": 0.3105948269367218, + "learning_rate": 9.992270339041123e-05, + "loss": 0.0658, + "step": 4010 + }, + { + "epoch": 0.12408173344033582, + "grad_norm": 0.33411526679992676, + "learning_rate": 9.992116405987053e-05, + "loss": 0.0695, + "step": 4020 + }, + { + "epoch": 0.12439039446879437, + "grad_norm": 0.3317610025405884, + "learning_rate": 9.991960956459828e-05, + "loss": 0.0691, + "step": 4030 + }, + { + "epoch": 0.12469905549725292, + "grad_norm": 0.2825261056423187, + "learning_rate": 9.991803990506669e-05, + "loss": 0.0621, + "step": 4040 + }, + { + "epoch": 0.12500771652571147, + "grad_norm": 0.33935701847076416, + "learning_rate": 9.991645508175258e-05, + "loss": 0.0588, + "step": 4050 + }, + { + "epoch": 0.12531637755417002, + "grad_norm": 0.5485620498657227, + "learning_rate": 9.99148550951374e-05, + "loss": 0.0667, + "step": 4060 + }, + { + "epoch": 0.12562503858262855, + "grad_norm": 0.25726214051246643, + "learning_rate": 9.991323994570716e-05, + "loss": 0.0643, + "step": 4070 + }, + { + "epoch": 0.1259336996110871, + "grad_norm": 0.34715521335601807, + "learning_rate": 9.99116096339525e-05, + "loss": 0.0646, + "step": 4080 + }, + { + "epoch": 0.12624236063954566, + "grad_norm": 0.25140416622161865, + "learning_rate": 9.990996416036869e-05, + "loss": 0.0672, + "step": 4090 + }, + { + "epoch": 0.1265510216680042, + "grad_norm": 0.3400493860244751, + "learning_rate": 9.990830352545555e-05, + "loss": 0.0613, + "step": 4100 + }, + { + "epoch": 0.12685968269646275, + "grad_norm": 0.20127159357070923, + "learning_rate": 9.990662772971756e-05, + "loss": 0.0677, + "step": 4110 + }, + { + "epoch": 0.12716834372492128, + "grad_norm": 0.2744612395763397, + "learning_rate": 9.990493677366376e-05, + "loss": 0.0658, + "step": 4120 + }, + { + "epoch": 0.12747700475337984, + "grad_norm": 0.3656170666217804, + "learning_rate": 9.990323065780786e-05, + "loss": 0.0641, + "step": 4130 + }, + { + "epoch": 0.1277856657818384, + "grad_norm": 0.3292602002620697, + "learning_rate": 9.990150938266808e-05, + "loss": 0.0707, + "step": 4140 + }, + { + "epoch": 0.12809432681029692, + "grad_norm": 0.32746320962905884, + "learning_rate": 9.989977294876733e-05, + "loss": 0.0612, + "step": 4150 + }, + { + "epoch": 0.12840298783875548, + "grad_norm": 0.3215785026550293, + "learning_rate": 9.989802135663308e-05, + "loss": 0.0735, + "step": 4160 + }, + { + "epoch": 0.12871164886721403, + "grad_norm": 0.42360180616378784, + "learning_rate": 9.989625460679743e-05, + "loss": 0.0634, + "step": 4170 + }, + { + "epoch": 0.12902030989567256, + "grad_norm": 0.5451992154121399, + "learning_rate": 9.989447269979706e-05, + "loss": 0.0678, + "step": 4180 + }, + { + "epoch": 0.12932897092413112, + "grad_norm": 0.3341560363769531, + "learning_rate": 9.989267563617328e-05, + "loss": 0.0628, + "step": 4190 + }, + { + "epoch": 0.12963763195258968, + "grad_norm": 0.3492256999015808, + "learning_rate": 9.989086341647198e-05, + "loss": 0.0634, + "step": 4200 + }, + { + "epoch": 0.1299462929810482, + "grad_norm": 0.23969393968582153, + "learning_rate": 9.988903604124366e-05, + "loss": 0.0612, + "step": 4210 + }, + { + "epoch": 0.13025495400950676, + "grad_norm": 0.3068995177745819, + "learning_rate": 9.988719351104343e-05, + "loss": 0.0593, + "step": 4220 + }, + { + "epoch": 0.13056361503796532, + "grad_norm": 0.3316771686077118, + "learning_rate": 9.9885335826431e-05, + "loss": 0.0622, + "step": 4230 + }, + { + "epoch": 0.13087227606642385, + "grad_norm": 0.33145400881767273, + "learning_rate": 9.988346298797071e-05, + "loss": 0.067, + "step": 4240 + }, + { + "epoch": 0.1311809370948824, + "grad_norm": 0.32665348052978516, + "learning_rate": 9.988157499623146e-05, + "loss": 0.0687, + "step": 4250 + }, + { + "epoch": 0.13148959812334093, + "grad_norm": 0.3688134253025055, + "learning_rate": 9.987967185178677e-05, + "loss": 0.065, + "step": 4260 + }, + { + "epoch": 0.1317982591517995, + "grad_norm": 0.2717166543006897, + "learning_rate": 9.987775355521476e-05, + "loss": 0.0599, + "step": 4270 + }, + { + "epoch": 0.13210692018025805, + "grad_norm": 0.33211439847946167, + "learning_rate": 9.987582010709817e-05, + "loss": 0.0661, + "step": 4280 + }, + { + "epoch": 0.13241558120871658, + "grad_norm": 0.3264717161655426, + "learning_rate": 9.987387150802431e-05, + "loss": 0.0645, + "step": 4290 + }, + { + "epoch": 0.13272424223717513, + "grad_norm": 0.3397989273071289, + "learning_rate": 9.987190775858517e-05, + "loss": 0.0663, + "step": 4300 + }, + { + "epoch": 0.1330329032656337, + "grad_norm": 0.35029137134552, + "learning_rate": 9.98699288593772e-05, + "loss": 0.066, + "step": 4310 + }, + { + "epoch": 0.13334156429409222, + "grad_norm": 0.27995622158050537, + "learning_rate": 9.986793481100161e-05, + "loss": 0.0633, + "step": 4320 + }, + { + "epoch": 0.13365022532255078, + "grad_norm": 0.33588993549346924, + "learning_rate": 9.986592561406412e-05, + "loss": 0.0644, + "step": 4330 + }, + { + "epoch": 0.13395888635100933, + "grad_norm": 0.29516837000846863, + "learning_rate": 9.986390126917503e-05, + "loss": 0.0673, + "step": 4340 + }, + { + "epoch": 0.13426754737946786, + "grad_norm": 0.2717601954936981, + "learning_rate": 9.986186177694933e-05, + "loss": 0.0641, + "step": 4350 + }, + { + "epoch": 0.13457620840792642, + "grad_norm": 0.280254602432251, + "learning_rate": 9.985980713800656e-05, + "loss": 0.0652, + "step": 4360 + }, + { + "epoch": 0.13488486943638497, + "grad_norm": 0.3060702979564667, + "learning_rate": 9.985773735297084e-05, + "loss": 0.0658, + "step": 4370 + }, + { + "epoch": 0.1351935304648435, + "grad_norm": 0.39383813738822937, + "learning_rate": 9.985565242247092e-05, + "loss": 0.0674, + "step": 4380 + }, + { + "epoch": 0.13550219149330206, + "grad_norm": 0.3188556432723999, + "learning_rate": 9.985355234714016e-05, + "loss": 0.0688, + "step": 4390 + }, + { + "epoch": 0.1358108525217606, + "grad_norm": 0.3824116587638855, + "learning_rate": 9.985143712761652e-05, + "loss": 0.0703, + "step": 4400 + }, + { + "epoch": 0.13611951355021915, + "grad_norm": 0.31102100014686584, + "learning_rate": 9.984930676454252e-05, + "loss": 0.0686, + "step": 4410 + }, + { + "epoch": 0.1364281745786777, + "grad_norm": 0.2906337380409241, + "learning_rate": 9.984716125856532e-05, + "loss": 0.0612, + "step": 4420 + }, + { + "epoch": 0.13673683560713623, + "grad_norm": 0.2502768635749817, + "learning_rate": 9.984500061033667e-05, + "loss": 0.0659, + "step": 4430 + }, + { + "epoch": 0.1370454966355948, + "grad_norm": 0.3983260989189148, + "learning_rate": 9.984282482051293e-05, + "loss": 0.0617, + "step": 4440 + }, + { + "epoch": 0.13735415766405334, + "grad_norm": 0.31727829575538635, + "learning_rate": 9.9840633889755e-05, + "loss": 0.0638, + "step": 4450 + }, + { + "epoch": 0.13766281869251187, + "grad_norm": 0.30548444390296936, + "learning_rate": 9.983842781872848e-05, + "loss": 0.064, + "step": 4460 + }, + { + "epoch": 0.13797147972097043, + "grad_norm": 0.37299174070358276, + "learning_rate": 9.98362066081035e-05, + "loss": 0.0655, + "step": 4470 + }, + { + "epoch": 0.138280140749429, + "grad_norm": 0.31549781560897827, + "learning_rate": 9.983397025855479e-05, + "loss": 0.063, + "step": 4480 + }, + { + "epoch": 0.13858880177788752, + "grad_norm": 0.6189848184585571, + "learning_rate": 9.983171877076171e-05, + "loss": 0.0662, + "step": 4490 + }, + { + "epoch": 0.13889746280634607, + "grad_norm": 0.3214423954486847, + "learning_rate": 9.98294521454082e-05, + "loss": 0.0668, + "step": 4500 + }, + { + "epoch": 0.13920612383480463, + "grad_norm": 0.2896527647972107, + "learning_rate": 9.98271703831828e-05, + "loss": 0.0637, + "step": 4510 + }, + { + "epoch": 0.13951478486326316, + "grad_norm": 0.3426837921142578, + "learning_rate": 9.982487348477865e-05, + "loss": 0.0597, + "step": 4520 + }, + { + "epoch": 0.13982344589172171, + "grad_norm": 0.37789902091026306, + "learning_rate": 9.982256145089347e-05, + "loss": 0.062, + "step": 4530 + }, + { + "epoch": 0.14013210692018027, + "grad_norm": 0.2660737633705139, + "learning_rate": 9.982023428222962e-05, + "loss": 0.0644, + "step": 4540 + }, + { + "epoch": 0.1404407679486388, + "grad_norm": 0.32760822772979736, + "learning_rate": 9.981789197949403e-05, + "loss": 0.0601, + "step": 4550 + }, + { + "epoch": 0.14074942897709736, + "grad_norm": 0.2989334762096405, + "learning_rate": 9.98155345433982e-05, + "loss": 0.0669, + "step": 4560 + }, + { + "epoch": 0.1410580900055559, + "grad_norm": 0.28945350646972656, + "learning_rate": 9.981316197465831e-05, + "loss": 0.0612, + "step": 4570 + }, + { + "epoch": 0.14136675103401444, + "grad_norm": 0.29931095242500305, + "learning_rate": 9.981077427399504e-05, + "loss": 0.0653, + "step": 4580 + }, + { + "epoch": 0.141675412062473, + "grad_norm": 0.3459862768650055, + "learning_rate": 9.980837144213371e-05, + "loss": 0.0676, + "step": 4590 + }, + { + "epoch": 0.14198407309093153, + "grad_norm": 0.36630016565322876, + "learning_rate": 9.980595347980426e-05, + "loss": 0.0656, + "step": 4600 + }, + { + "epoch": 0.14229273411939009, + "grad_norm": 0.3308151066303253, + "learning_rate": 9.980352038774119e-05, + "loss": 0.0639, + "step": 4610 + }, + { + "epoch": 0.14260139514784864, + "grad_norm": 0.3080301880836487, + "learning_rate": 9.98010721666836e-05, + "loss": 0.0615, + "step": 4620 + }, + { + "epoch": 0.14291005617630717, + "grad_norm": 0.3822731077671051, + "learning_rate": 9.979860881737523e-05, + "loss": 0.0641, + "step": 4630 + }, + { + "epoch": 0.14321871720476573, + "grad_norm": 0.8306577801704407, + "learning_rate": 9.979613034056434e-05, + "loss": 0.0631, + "step": 4640 + }, + { + "epoch": 0.14352737823322428, + "grad_norm": 0.3764645755290985, + "learning_rate": 9.979363673700386e-05, + "loss": 0.0641, + "step": 4650 + }, + { + "epoch": 0.1438360392616828, + "grad_norm": 0.2888333797454834, + "learning_rate": 9.979112800745124e-05, + "loss": 0.0616, + "step": 4660 + }, + { + "epoch": 0.14414470029014137, + "grad_norm": 0.32029491662979126, + "learning_rate": 9.978860415266861e-05, + "loss": 0.0648, + "step": 4670 + }, + { + "epoch": 0.14445336131859993, + "grad_norm": 0.3063207268714905, + "learning_rate": 9.978606517342262e-05, + "loss": 0.0647, + "step": 4680 + }, + { + "epoch": 0.14476202234705846, + "grad_norm": 0.27233627438545227, + "learning_rate": 9.978351107048456e-05, + "loss": 0.0691, + "step": 4690 + }, + { + "epoch": 0.145070683375517, + "grad_norm": 0.2633320689201355, + "learning_rate": 9.978094184463029e-05, + "loss": 0.0627, + "step": 4700 + }, + { + "epoch": 0.14537934440397554, + "grad_norm": 0.31424203515052795, + "learning_rate": 9.977835749664029e-05, + "loss": 0.065, + "step": 4710 + }, + { + "epoch": 0.1456880054324341, + "grad_norm": 0.3194933533668518, + "learning_rate": 9.97757580272996e-05, + "loss": 0.0647, + "step": 4720 + }, + { + "epoch": 0.14599666646089265, + "grad_norm": 0.30919909477233887, + "learning_rate": 9.977314343739786e-05, + "loss": 0.0634, + "step": 4730 + }, + { + "epoch": 0.14630532748935118, + "grad_norm": 0.3083547055721283, + "learning_rate": 9.977051372772934e-05, + "loss": 0.064, + "step": 4740 + }, + { + "epoch": 0.14661398851780974, + "grad_norm": 0.3089055120944977, + "learning_rate": 9.976786889909286e-05, + "loss": 0.0672, + "step": 4750 + }, + { + "epoch": 0.1469226495462683, + "grad_norm": 0.3202751874923706, + "learning_rate": 9.976520895229185e-05, + "loss": 0.0675, + "step": 4760 + }, + { + "epoch": 0.14723131057472683, + "grad_norm": 0.32721221446990967, + "learning_rate": 9.976253388813433e-05, + "loss": 0.0676, + "step": 4770 + }, + { + "epoch": 0.14753997160318538, + "grad_norm": 0.46344634890556335, + "learning_rate": 9.975984370743293e-05, + "loss": 0.0701, + "step": 4780 + }, + { + "epoch": 0.14784863263164394, + "grad_norm": 0.24959151446819305, + "learning_rate": 9.975713841100485e-05, + "loss": 0.0654, + "step": 4790 + }, + { + "epoch": 0.14815729366010247, + "grad_norm": 0.3499682545661926, + "learning_rate": 9.975441799967187e-05, + "loss": 0.062, + "step": 4800 + }, + { + "epoch": 0.14846595468856102, + "grad_norm": 0.2373305857181549, + "learning_rate": 9.975168247426039e-05, + "loss": 0.0713, + "step": 4810 + }, + { + "epoch": 0.14877461571701958, + "grad_norm": 0.32554590702056885, + "learning_rate": 9.974893183560139e-05, + "loss": 0.0627, + "step": 4820 + }, + { + "epoch": 0.1490832767454781, + "grad_norm": 0.2868477702140808, + "learning_rate": 9.974616608453045e-05, + "loss": 0.0664, + "step": 4830 + }, + { + "epoch": 0.14939193777393667, + "grad_norm": 0.3110666275024414, + "learning_rate": 9.974338522188772e-05, + "loss": 0.0656, + "step": 4840 + }, + { + "epoch": 0.1497005988023952, + "grad_norm": 0.37541258335113525, + "learning_rate": 9.974058924851797e-05, + "loss": 0.0624, + "step": 4850 + }, + { + "epoch": 0.15000925983085375, + "grad_norm": 0.31470367312431335, + "learning_rate": 9.973777816527051e-05, + "loss": 0.0643, + "step": 4860 + }, + { + "epoch": 0.1503179208593123, + "grad_norm": 0.3141568899154663, + "learning_rate": 9.973495197299931e-05, + "loss": 0.0617, + "step": 4870 + }, + { + "epoch": 0.15062658188777084, + "grad_norm": 0.22712741792201996, + "learning_rate": 9.973211067256287e-05, + "loss": 0.0653, + "step": 4880 + }, + { + "epoch": 0.1509352429162294, + "grad_norm": 0.4370805025100708, + "learning_rate": 9.97292542648243e-05, + "loss": 0.0639, + "step": 4890 + }, + { + "epoch": 0.15124390394468795, + "grad_norm": 0.3022097945213318, + "learning_rate": 9.972638275065131e-05, + "loss": 0.0624, + "step": 4900 + }, + { + "epoch": 0.15155256497314648, + "grad_norm": 0.24599198997020721, + "learning_rate": 9.972349613091621e-05, + "loss": 0.0617, + "step": 4910 + }, + { + "epoch": 0.15186122600160504, + "grad_norm": 0.2811563014984131, + "learning_rate": 9.972059440649584e-05, + "loss": 0.069, + "step": 4920 + }, + { + "epoch": 0.1521698870300636, + "grad_norm": 0.3346795439720154, + "learning_rate": 9.971767757827168e-05, + "loss": 0.0656, + "step": 4930 + }, + { + "epoch": 0.15247854805852212, + "grad_norm": 0.46872568130493164, + "learning_rate": 9.971474564712982e-05, + "loss": 0.0618, + "step": 4940 + }, + { + "epoch": 0.15278720908698068, + "grad_norm": 0.2738671898841858, + "learning_rate": 9.971179861396084e-05, + "loss": 0.064, + "step": 4950 + }, + { + "epoch": 0.15309587011543924, + "grad_norm": 0.27366095781326294, + "learning_rate": 9.970883647966003e-05, + "loss": 0.0646, + "step": 4960 + }, + { + "epoch": 0.15340453114389777, + "grad_norm": 0.3603101968765259, + "learning_rate": 9.970585924512717e-05, + "loss": 0.0635, + "step": 4970 + }, + { + "epoch": 0.15371319217235632, + "grad_norm": 0.33699992299079895, + "learning_rate": 9.970286691126669e-05, + "loss": 0.0665, + "step": 4980 + }, + { + "epoch": 0.15402185320081488, + "grad_norm": 0.3092288076877594, + "learning_rate": 9.969985947898756e-05, + "loss": 0.0639, + "step": 4990 + }, + { + "epoch": 0.1543305142292734, + "grad_norm": 0.312682181596756, + "learning_rate": 9.969683694920337e-05, + "loss": 0.0667, + "step": 5000 + }, + { + "epoch": 0.15463917525773196, + "grad_norm": 0.31372886896133423, + "learning_rate": 9.969379932283228e-05, + "loss": 0.0638, + "step": 5010 + }, + { + "epoch": 0.1549478362861905, + "grad_norm": 0.34410881996154785, + "learning_rate": 9.969074660079704e-05, + "loss": 0.0666, + "step": 5020 + }, + { + "epoch": 0.15525649731464905, + "grad_norm": 0.36136388778686523, + "learning_rate": 9.968767878402501e-05, + "loss": 0.0688, + "step": 5030 + }, + { + "epoch": 0.1555651583431076, + "grad_norm": 0.24147990345954895, + "learning_rate": 9.968459587344808e-05, + "loss": 0.0632, + "step": 5040 + }, + { + "epoch": 0.15587381937156614, + "grad_norm": 0.5412231087684631, + "learning_rate": 9.968149787000278e-05, + "loss": 0.0607, + "step": 5050 + }, + { + "epoch": 0.1561824804000247, + "grad_norm": 0.2927287220954895, + "learning_rate": 9.967838477463018e-05, + "loss": 0.0624, + "step": 5060 + }, + { + "epoch": 0.15649114142848325, + "grad_norm": 0.3407946527004242, + "learning_rate": 9.967525658827597e-05, + "loss": 0.0624, + "step": 5070 + }, + { + "epoch": 0.15679980245694178, + "grad_norm": 0.5399467349052429, + "learning_rate": 9.967211331189042e-05, + "loss": 0.06, + "step": 5080 + }, + { + "epoch": 0.15710846348540033, + "grad_norm": 0.2979530096054077, + "learning_rate": 9.966895494642834e-05, + "loss": 0.0672, + "step": 5090 + }, + { + "epoch": 0.1574171245138589, + "grad_norm": 0.26906707882881165, + "learning_rate": 9.96657814928492e-05, + "loss": 0.0636, + "step": 5100 + }, + { + "epoch": 0.15772578554231742, + "grad_norm": 0.29622334241867065, + "learning_rate": 9.966259295211697e-05, + "loss": 0.0617, + "step": 5110 + }, + { + "epoch": 0.15803444657077598, + "grad_norm": 0.40160709619522095, + "learning_rate": 9.965938932520028e-05, + "loss": 0.0637, + "step": 5120 + }, + { + "epoch": 0.15834310759923453, + "grad_norm": 0.2675601541996002, + "learning_rate": 9.965617061307229e-05, + "loss": 0.0606, + "step": 5130 + }, + { + "epoch": 0.15865176862769306, + "grad_norm": 0.287625253200531, + "learning_rate": 9.965293681671077e-05, + "loss": 0.0683, + "step": 5140 + }, + { + "epoch": 0.15896042965615162, + "grad_norm": 0.3972514867782593, + "learning_rate": 9.964968793709804e-05, + "loss": 0.064, + "step": 5150 + }, + { + "epoch": 0.15926909068461015, + "grad_norm": 0.24673287570476532, + "learning_rate": 9.964642397522106e-05, + "loss": 0.0621, + "step": 5160 + }, + { + "epoch": 0.1595777517130687, + "grad_norm": 0.344125360250473, + "learning_rate": 9.96431449320713e-05, + "loss": 0.062, + "step": 5170 + }, + { + "epoch": 0.15988641274152726, + "grad_norm": 0.40750396251678467, + "learning_rate": 9.963985080864486e-05, + "loss": 0.0584, + "step": 5180 + }, + { + "epoch": 0.1601950737699858, + "grad_norm": 0.3539929986000061, + "learning_rate": 9.96365416059424e-05, + "loss": 0.0703, + "step": 5190 + }, + { + "epoch": 0.16050373479844435, + "grad_norm": 0.3243013322353363, + "learning_rate": 9.963321732496919e-05, + "loss": 0.0623, + "step": 5200 + }, + { + "epoch": 0.1608123958269029, + "grad_norm": 0.3342839479446411, + "learning_rate": 9.962987796673506e-05, + "loss": 0.0599, + "step": 5210 + }, + { + "epoch": 0.16112105685536143, + "grad_norm": 0.30589887499809265, + "learning_rate": 9.962652353225438e-05, + "loss": 0.0662, + "step": 5220 + }, + { + "epoch": 0.16142971788382, + "grad_norm": 0.3918161988258362, + "learning_rate": 9.962315402254619e-05, + "loss": 0.0631, + "step": 5230 + }, + { + "epoch": 0.16173837891227855, + "grad_norm": 2.7521450519561768, + "learning_rate": 9.9619769438634e-05, + "loss": 0.0634, + "step": 5240 + }, + { + "epoch": 0.16204703994073708, + "grad_norm": 0.34996548295021057, + "learning_rate": 9.9616369781546e-05, + "loss": 0.0638, + "step": 5250 + }, + { + "epoch": 0.16235570096919563, + "grad_norm": 0.3157466948032379, + "learning_rate": 9.961295505231491e-05, + "loss": 0.0655, + "step": 5260 + }, + { + "epoch": 0.1626643619976542, + "grad_norm": 0.3169466555118561, + "learning_rate": 9.960952525197804e-05, + "loss": 0.0673, + "step": 5270 + }, + { + "epoch": 0.16297302302611272, + "grad_norm": 0.2807951867580414, + "learning_rate": 9.960608038157724e-05, + "loss": 0.0638, + "step": 5280 + }, + { + "epoch": 0.16328168405457127, + "grad_norm": 0.3525990843772888, + "learning_rate": 9.960262044215901e-05, + "loss": 0.0657, + "step": 5290 + }, + { + "epoch": 0.1635903450830298, + "grad_norm": 0.4704975187778473, + "learning_rate": 9.959914543477435e-05, + "loss": 0.0618, + "step": 5300 + }, + { + "epoch": 0.16389900611148836, + "grad_norm": 0.32489126920700073, + "learning_rate": 9.959565536047892e-05, + "loss": 0.0639, + "step": 5310 + }, + { + "epoch": 0.16420766713994692, + "grad_norm": 0.5247007012367249, + "learning_rate": 9.959215022033288e-05, + "loss": 0.0601, + "step": 5320 + }, + { + "epoch": 0.16451632816840545, + "grad_norm": 0.22159357368946075, + "learning_rate": 9.9588630015401e-05, + "loss": 0.06, + "step": 5330 + }, + { + "epoch": 0.164824989196864, + "grad_norm": 0.30226215720176697, + "learning_rate": 9.958509474675264e-05, + "loss": 0.0645, + "step": 5340 + }, + { + "epoch": 0.16513365022532256, + "grad_norm": 0.28219372034072876, + "learning_rate": 9.958154441546171e-05, + "loss": 0.0634, + "step": 5350 + }, + { + "epoch": 0.1654423112537811, + "grad_norm": 0.2935613691806793, + "learning_rate": 9.957797902260673e-05, + "loss": 0.0602, + "step": 5360 + }, + { + "epoch": 0.16575097228223964, + "grad_norm": 0.36106252670288086, + "learning_rate": 9.957439856927073e-05, + "loss": 0.0642, + "step": 5370 + }, + { + "epoch": 0.1660596333106982, + "grad_norm": 0.2979830205440521, + "learning_rate": 9.957080305654139e-05, + "loss": 0.0621, + "step": 5380 + }, + { + "epoch": 0.16636829433915673, + "grad_norm": 0.21104472875595093, + "learning_rate": 9.956719248551092e-05, + "loss": 0.0611, + "step": 5390 + }, + { + "epoch": 0.1666769553676153, + "grad_norm": 0.3817431330680847, + "learning_rate": 9.956356685727612e-05, + "loss": 0.0616, + "step": 5400 + }, + { + "epoch": 0.16698561639607384, + "grad_norm": 0.33386072516441345, + "learning_rate": 9.955992617293836e-05, + "loss": 0.0651, + "step": 5410 + }, + { + "epoch": 0.16729427742453237, + "grad_norm": 0.2616725265979767, + "learning_rate": 9.955627043360358e-05, + "loss": 0.0654, + "step": 5420 + }, + { + "epoch": 0.16760293845299093, + "grad_norm": 0.26916608214378357, + "learning_rate": 9.955259964038231e-05, + "loss": 0.0593, + "step": 5430 + }, + { + "epoch": 0.16791159948144946, + "grad_norm": 0.31779879331588745, + "learning_rate": 9.954891379438962e-05, + "loss": 0.059, + "step": 5440 + }, + { + "epoch": 0.16822026050990802, + "grad_norm": 0.3522229790687561, + "learning_rate": 9.954521289674519e-05, + "loss": 0.0637, + "step": 5450 + }, + { + "epoch": 0.16852892153836657, + "grad_norm": 0.34345704317092896, + "learning_rate": 9.954149694857325e-05, + "loss": 0.0568, + "step": 5460 + }, + { + "epoch": 0.1688375825668251, + "grad_norm": 0.3728286325931549, + "learning_rate": 9.953776595100258e-05, + "loss": 0.0604, + "step": 5470 + }, + { + "epoch": 0.16914624359528366, + "grad_norm": 0.28294381499290466, + "learning_rate": 9.95340199051666e-05, + "loss": 0.0616, + "step": 5480 + }, + { + "epoch": 0.16945490462374221, + "grad_norm": 0.3584555983543396, + "learning_rate": 9.953025881220325e-05, + "loss": 0.0615, + "step": 5490 + }, + { + "epoch": 0.16976356565220074, + "grad_norm": 0.3572008013725281, + "learning_rate": 9.952648267325504e-05, + "loss": 0.0676, + "step": 5500 + }, + { + "epoch": 0.1700722266806593, + "grad_norm": 0.2999359369277954, + "learning_rate": 9.952269148946905e-05, + "loss": 0.0626, + "step": 5510 + }, + { + "epoch": 0.17038088770911786, + "grad_norm": 0.46514245867729187, + "learning_rate": 9.951888526199697e-05, + "loss": 0.0687, + "step": 5520 + }, + { + "epoch": 0.17068954873757639, + "grad_norm": 0.30704766511917114, + "learning_rate": 9.951506399199501e-05, + "loss": 0.0565, + "step": 5530 + }, + { + "epoch": 0.17099820976603494, + "grad_norm": 0.21310271322727203, + "learning_rate": 9.951122768062399e-05, + "loss": 0.0661, + "step": 5540 + }, + { + "epoch": 0.1713068707944935, + "grad_norm": 0.30999985337257385, + "learning_rate": 9.950737632904927e-05, + "loss": 0.0633, + "step": 5550 + }, + { + "epoch": 0.17161553182295203, + "grad_norm": 0.31819024682044983, + "learning_rate": 9.950350993844077e-05, + "loss": 0.0656, + "step": 5560 + }, + { + "epoch": 0.17192419285141058, + "grad_norm": 0.3773069679737091, + "learning_rate": 9.949962850997303e-05, + "loss": 0.0633, + "step": 5570 + }, + { + "epoch": 0.17223285387986914, + "grad_norm": 0.4892944097518921, + "learning_rate": 9.949573204482512e-05, + "loss": 0.07, + "step": 5580 + }, + { + "epoch": 0.17254151490832767, + "grad_norm": 0.2802511155605316, + "learning_rate": 9.949182054418064e-05, + "loss": 0.065, + "step": 5590 + }, + { + "epoch": 0.17285017593678623, + "grad_norm": 0.2616889476776123, + "learning_rate": 9.948789400922787e-05, + "loss": 0.0683, + "step": 5600 + }, + { + "epoch": 0.17315883696524476, + "grad_norm": 0.3976101577281952, + "learning_rate": 9.948395244115953e-05, + "loss": 0.0692, + "step": 5610 + }, + { + "epoch": 0.1734674979937033, + "grad_norm": 0.35479703545570374, + "learning_rate": 9.9479995841173e-05, + "loss": 0.0637, + "step": 5620 + }, + { + "epoch": 0.17377615902216187, + "grad_norm": 0.36311113834381104, + "learning_rate": 9.947602421047017e-05, + "loss": 0.0655, + "step": 5630 + }, + { + "epoch": 0.1740848200506204, + "grad_norm": 0.2893132269382477, + "learning_rate": 9.947203755025753e-05, + "loss": 0.0614, + "step": 5640 + }, + { + "epoch": 0.17439348107907895, + "grad_norm": 0.21586774289608002, + "learning_rate": 9.946803586174611e-05, + "loss": 0.0667, + "step": 5650 + }, + { + "epoch": 0.1747021421075375, + "grad_norm": 0.3045802116394043, + "learning_rate": 9.946401914615151e-05, + "loss": 0.0602, + "step": 5660 + }, + { + "epoch": 0.17501080313599604, + "grad_norm": 0.24822203814983368, + "learning_rate": 9.945998740469394e-05, + "loss": 0.0633, + "step": 5670 + }, + { + "epoch": 0.1753194641644546, + "grad_norm": 0.31332525610923767, + "learning_rate": 9.945594063859809e-05, + "loss": 0.0591, + "step": 5680 + }, + { + "epoch": 0.17562812519291315, + "grad_norm": 0.5208169221878052, + "learning_rate": 9.94518788490933e-05, + "loss": 0.065, + "step": 5690 + }, + { + "epoch": 0.17593678622137168, + "grad_norm": 0.28817427158355713, + "learning_rate": 9.944780203741341e-05, + "loss": 0.0527, + "step": 5700 + }, + { + "epoch": 0.17624544724983024, + "grad_norm": 0.3200453817844391, + "learning_rate": 9.944371020479686e-05, + "loss": 0.0656, + "step": 5710 + }, + { + "epoch": 0.1765541082782888, + "grad_norm": 0.2970149517059326, + "learning_rate": 9.943960335248662e-05, + "loss": 0.0639, + "step": 5720 + }, + { + "epoch": 0.17686276930674732, + "grad_norm": 0.3014604449272156, + "learning_rate": 9.943548148173027e-05, + "loss": 0.068, + "step": 5730 + }, + { + "epoch": 0.17717143033520588, + "grad_norm": 0.27675285935401917, + "learning_rate": 9.943134459377992e-05, + "loss": 0.0655, + "step": 5740 + }, + { + "epoch": 0.1774800913636644, + "grad_norm": 0.3159029483795166, + "learning_rate": 9.942719268989222e-05, + "loss": 0.0591, + "step": 5750 + }, + { + "epoch": 0.17778875239212297, + "grad_norm": 0.40833351016044617, + "learning_rate": 9.942302577132844e-05, + "loss": 0.0605, + "step": 5760 + }, + { + "epoch": 0.17809741342058152, + "grad_norm": 0.30146297812461853, + "learning_rate": 9.941884383935438e-05, + "loss": 0.0598, + "step": 5770 + }, + { + "epoch": 0.17840607444904005, + "grad_norm": 0.3090271055698395, + "learning_rate": 9.941464689524039e-05, + "loss": 0.0642, + "step": 5780 + }, + { + "epoch": 0.1787147354774986, + "grad_norm": 0.22740520536899567, + "learning_rate": 9.941043494026139e-05, + "loss": 0.0582, + "step": 5790 + }, + { + "epoch": 0.17902339650595717, + "grad_norm": 0.3122260272502899, + "learning_rate": 9.940620797569685e-05, + "loss": 0.0641, + "step": 5800 + }, + { + "epoch": 0.1793320575344157, + "grad_norm": 0.2863333821296692, + "learning_rate": 9.940196600283082e-05, + "loss": 0.0637, + "step": 5810 + }, + { + "epoch": 0.17964071856287425, + "grad_norm": 0.38296836614608765, + "learning_rate": 9.939770902295192e-05, + "loss": 0.0627, + "step": 5820 + }, + { + "epoch": 0.1799493795913328, + "grad_norm": 0.288585364818573, + "learning_rate": 9.939343703735329e-05, + "loss": 0.0644, + "step": 5830 + }, + { + "epoch": 0.18025804061979134, + "grad_norm": 0.4764525592327118, + "learning_rate": 9.938915004733264e-05, + "loss": 0.0634, + "step": 5840 + }, + { + "epoch": 0.1805667016482499, + "grad_norm": 0.2876056730747223, + "learning_rate": 9.938484805419224e-05, + "loss": 0.0633, + "step": 5850 + }, + { + "epoch": 0.18087536267670845, + "grad_norm": 0.3649607002735138, + "learning_rate": 9.938053105923894e-05, + "loss": 0.061, + "step": 5860 + }, + { + "epoch": 0.18118402370516698, + "grad_norm": 0.42840108275413513, + "learning_rate": 9.937619906378413e-05, + "loss": 0.0606, + "step": 5870 + }, + { + "epoch": 0.18149268473362554, + "grad_norm": 0.3018542528152466, + "learning_rate": 9.937185206914374e-05, + "loss": 0.0583, + "step": 5880 + }, + { + "epoch": 0.18180134576208407, + "grad_norm": 0.2915003299713135, + "learning_rate": 9.936749007663829e-05, + "loss": 0.0551, + "step": 5890 + }, + { + "epoch": 0.18211000679054262, + "grad_norm": 0.3344627916812897, + "learning_rate": 9.93631130875928e-05, + "loss": 0.0605, + "step": 5900 + }, + { + "epoch": 0.18241866781900118, + "grad_norm": 0.3326474130153656, + "learning_rate": 9.935872110333692e-05, + "loss": 0.0627, + "step": 5910 + }, + { + "epoch": 0.1827273288474597, + "grad_norm": 0.25983861088752747, + "learning_rate": 9.935431412520484e-05, + "loss": 0.063, + "step": 5920 + }, + { + "epoch": 0.18303598987591826, + "grad_norm": 0.48817339539527893, + "learning_rate": 9.934989215453523e-05, + "loss": 0.0679, + "step": 5930 + }, + { + "epoch": 0.18334465090437682, + "grad_norm": 0.40757983922958374, + "learning_rate": 9.934545519267139e-05, + "loss": 0.0638, + "step": 5940 + }, + { + "epoch": 0.18365331193283535, + "grad_norm": 0.33245670795440674, + "learning_rate": 9.934100324096117e-05, + "loss": 0.0634, + "step": 5950 + }, + { + "epoch": 0.1839619729612939, + "grad_norm": 0.30703914165496826, + "learning_rate": 9.933653630075692e-05, + "loss": 0.0622, + "step": 5960 + }, + { + "epoch": 0.18427063398975246, + "grad_norm": 0.5176114439964294, + "learning_rate": 9.93320543734156e-05, + "loss": 0.0604, + "step": 5970 + }, + { + "epoch": 0.184579295018211, + "grad_norm": 0.433695912361145, + "learning_rate": 9.932755746029871e-05, + "loss": 0.0587, + "step": 5980 + }, + { + "epoch": 0.18488795604666955, + "grad_norm": 0.4775887131690979, + "learning_rate": 9.932304556277228e-05, + "loss": 0.0598, + "step": 5990 + }, + { + "epoch": 0.1851966170751281, + "grad_norm": 0.3619529902935028, + "learning_rate": 9.93185186822069e-05, + "loss": 0.0593, + "step": 6000 + }, + { + "epoch": 0.18550527810358663, + "grad_norm": 0.342631071805954, + "learning_rate": 9.931397681997773e-05, + "loss": 0.0623, + "step": 6010 + }, + { + "epoch": 0.1858139391320452, + "grad_norm": 0.2774757444858551, + "learning_rate": 9.930941997746446e-05, + "loss": 0.0601, + "step": 6020 + }, + { + "epoch": 0.18612260016050375, + "grad_norm": 0.3697463274002075, + "learning_rate": 9.930484815605134e-05, + "loss": 0.066, + "step": 6030 + }, + { + "epoch": 0.18643126118896228, + "grad_norm": 0.302310585975647, + "learning_rate": 9.930026135712717e-05, + "loss": 0.0607, + "step": 6040 + }, + { + "epoch": 0.18673992221742083, + "grad_norm": 0.4300357401371002, + "learning_rate": 9.92956595820853e-05, + "loss": 0.0589, + "step": 6050 + }, + { + "epoch": 0.18704858324587936, + "grad_norm": 0.24319635331630707, + "learning_rate": 9.929104283232362e-05, + "loss": 0.0618, + "step": 6060 + }, + { + "epoch": 0.18735724427433792, + "grad_norm": 0.32955580949783325, + "learning_rate": 9.92864111092446e-05, + "loss": 0.0635, + "step": 6070 + }, + { + "epoch": 0.18766590530279648, + "grad_norm": 0.26122012734413147, + "learning_rate": 9.92817644142552e-05, + "loss": 0.0671, + "step": 6080 + }, + { + "epoch": 0.187974566331255, + "grad_norm": 0.3200381100177765, + "learning_rate": 9.927710274876698e-05, + "loss": 0.0624, + "step": 6090 + }, + { + "epoch": 0.18828322735971356, + "grad_norm": 0.507836103439331, + "learning_rate": 9.927242611419603e-05, + "loss": 0.063, + "step": 6100 + }, + { + "epoch": 0.18859188838817212, + "grad_norm": 0.40854567289352417, + "learning_rate": 9.926773451196301e-05, + "loss": 0.0631, + "step": 6110 + }, + { + "epoch": 0.18890054941663065, + "grad_norm": 0.37172695994377136, + "learning_rate": 9.926302794349306e-05, + "loss": 0.0649, + "step": 6120 + }, + { + "epoch": 0.1892092104450892, + "grad_norm": 0.2771741449832916, + "learning_rate": 9.925830641021594e-05, + "loss": 0.0614, + "step": 6130 + }, + { + "epoch": 0.18951787147354776, + "grad_norm": 0.31410354375839233, + "learning_rate": 9.925356991356593e-05, + "loss": 0.0655, + "step": 6140 + }, + { + "epoch": 0.1898265325020063, + "grad_norm": 0.3322015106678009, + "learning_rate": 9.924881845498184e-05, + "loss": 0.0615, + "step": 6150 + }, + { + "epoch": 0.19013519353046485, + "grad_norm": 0.32345253229141235, + "learning_rate": 9.924405203590705e-05, + "loss": 0.0609, + "step": 6160 + }, + { + "epoch": 0.1904438545589234, + "grad_norm": 0.2657075822353363, + "learning_rate": 9.923927065778946e-05, + "loss": 0.0596, + "step": 6170 + }, + { + "epoch": 0.19075251558738193, + "grad_norm": 0.3078175485134125, + "learning_rate": 9.923447432208154e-05, + "loss": 0.0612, + "step": 6180 + }, + { + "epoch": 0.1910611766158405, + "grad_norm": 0.3500000834465027, + "learning_rate": 9.922966303024027e-05, + "loss": 0.0647, + "step": 6190 + }, + { + "epoch": 0.19136983764429902, + "grad_norm": 0.3014927804470062, + "learning_rate": 9.922483678372721e-05, + "loss": 0.0641, + "step": 6200 + }, + { + "epoch": 0.19167849867275757, + "grad_norm": 0.3278810679912567, + "learning_rate": 9.921999558400845e-05, + "loss": 0.067, + "step": 6210 + }, + { + "epoch": 0.19198715970121613, + "grad_norm": 0.3099312484264374, + "learning_rate": 9.92151394325546e-05, + "loss": 0.0623, + "step": 6220 + }, + { + "epoch": 0.19229582072967466, + "grad_norm": 0.3166517913341522, + "learning_rate": 9.921026833084084e-05, + "loss": 0.0619, + "step": 6230 + }, + { + "epoch": 0.19260448175813322, + "grad_norm": 0.3145199418067932, + "learning_rate": 9.920538228034689e-05, + "loss": 0.0621, + "step": 6240 + }, + { + "epoch": 0.19291314278659177, + "grad_norm": 0.3218872845172882, + "learning_rate": 9.920048128255699e-05, + "loss": 0.0613, + "step": 6250 + }, + { + "epoch": 0.1932218038150503, + "grad_norm": 0.3812282085418701, + "learning_rate": 9.919556533895995e-05, + "loss": 0.0574, + "step": 6260 + }, + { + "epoch": 0.19353046484350886, + "grad_norm": 0.364217072725296, + "learning_rate": 9.919063445104907e-05, + "loss": 0.0601, + "step": 6270 + }, + { + "epoch": 0.19383912587196742, + "grad_norm": 0.3083874583244324, + "learning_rate": 9.918568862032227e-05, + "loss": 0.058, + "step": 6280 + }, + { + "epoch": 0.19414778690042594, + "grad_norm": 0.35470545291900635, + "learning_rate": 9.918072784828194e-05, + "loss": 0.0647, + "step": 6290 + }, + { + "epoch": 0.1944564479288845, + "grad_norm": 0.6382593512535095, + "learning_rate": 9.917575213643501e-05, + "loss": 0.0613, + "step": 6300 + }, + { + "epoch": 0.19476510895734306, + "grad_norm": 0.3964184820652008, + "learning_rate": 9.917076148629302e-05, + "loss": 0.0605, + "step": 6310 + }, + { + "epoch": 0.1950737699858016, + "grad_norm": 0.2658218443393707, + "learning_rate": 9.916575589937196e-05, + "loss": 0.0624, + "step": 6320 + }, + { + "epoch": 0.19538243101426014, + "grad_norm": 0.2583714425563812, + "learning_rate": 9.916073537719239e-05, + "loss": 0.0603, + "step": 6330 + }, + { + "epoch": 0.19569109204271867, + "grad_norm": 0.32777664065361023, + "learning_rate": 9.915569992127944e-05, + "loss": 0.0616, + "step": 6340 + }, + { + "epoch": 0.19599975307117723, + "grad_norm": 0.2707041800022125, + "learning_rate": 9.915064953316273e-05, + "loss": 0.0697, + "step": 6350 + }, + { + "epoch": 0.1963084140996358, + "grad_norm": 0.4143170714378357, + "learning_rate": 9.914558421437645e-05, + "loss": 0.0578, + "step": 6360 + }, + { + "epoch": 0.19661707512809432, + "grad_norm": 0.32278159260749817, + "learning_rate": 9.914050396645929e-05, + "loss": 0.0614, + "step": 6370 + }, + { + "epoch": 0.19692573615655287, + "grad_norm": 0.30654376745224, + "learning_rate": 9.913540879095452e-05, + "loss": 0.0571, + "step": 6380 + }, + { + "epoch": 0.19723439718501143, + "grad_norm": 0.2563747763633728, + "learning_rate": 9.913029868940987e-05, + "loss": 0.0619, + "step": 6390 + }, + { + "epoch": 0.19754305821346996, + "grad_norm": 0.37831416726112366, + "learning_rate": 9.912517366337772e-05, + "loss": 0.0653, + "step": 6400 + }, + { + "epoch": 0.19785171924192851, + "grad_norm": 0.3160516619682312, + "learning_rate": 9.912003371441487e-05, + "loss": 0.0677, + "step": 6410 + }, + { + "epoch": 0.19816038027038707, + "grad_norm": 0.31447547674179077, + "learning_rate": 9.911487884408271e-05, + "loss": 0.0644, + "step": 6420 + }, + { + "epoch": 0.1984690412988456, + "grad_norm": 0.44800180196762085, + "learning_rate": 9.910970905394719e-05, + "loss": 0.0626, + "step": 6430 + }, + { + "epoch": 0.19877770232730416, + "grad_norm": 0.3292655050754547, + "learning_rate": 9.91045243455787e-05, + "loss": 0.0623, + "step": 6440 + }, + { + "epoch": 0.1990863633557627, + "grad_norm": 0.4155082404613495, + "learning_rate": 9.909932472055225e-05, + "loss": 0.0615, + "step": 6450 + }, + { + "epoch": 0.19939502438422124, + "grad_norm": 0.3022844195365906, + "learning_rate": 9.909411018044734e-05, + "loss": 0.0643, + "step": 6460 + }, + { + "epoch": 0.1997036854126798, + "grad_norm": 0.3439938426017761, + "learning_rate": 9.908888072684802e-05, + "loss": 0.0593, + "step": 6470 + }, + { + "epoch": 0.20001234644113836, + "grad_norm": 0.5130496621131897, + "learning_rate": 9.908363636134285e-05, + "loss": 0.0622, + "step": 6480 + }, + { + "epoch": 0.20032100746959688, + "grad_norm": 0.3896457552909851, + "learning_rate": 9.907837708552493e-05, + "loss": 0.0631, + "step": 6490 + }, + { + "epoch": 0.20062966849805544, + "grad_norm": 0.4565410017967224, + "learning_rate": 9.90731029009919e-05, + "loss": 0.0629, + "step": 6500 + }, + { + "epoch": 0.20093832952651397, + "grad_norm": 0.36794620752334595, + "learning_rate": 9.906781380934589e-05, + "loss": 0.068, + "step": 6510 + }, + { + "epoch": 0.20124699055497253, + "grad_norm": 0.3289928734302521, + "learning_rate": 9.906250981219362e-05, + "loss": 0.0587, + "step": 6520 + }, + { + "epoch": 0.20155565158343108, + "grad_norm": 0.4308786988258362, + "learning_rate": 9.905719091114628e-05, + "loss": 0.0644, + "step": 6530 + }, + { + "epoch": 0.2018643126118896, + "grad_norm": 0.4014663100242615, + "learning_rate": 9.905185710781964e-05, + "loss": 0.0576, + "step": 6540 + }, + { + "epoch": 0.20217297364034817, + "grad_norm": 0.32096153497695923, + "learning_rate": 9.904650840383392e-05, + "loss": 0.066, + "step": 6550 + }, + { + "epoch": 0.20248163466880673, + "grad_norm": 0.3060157001018524, + "learning_rate": 9.904114480081397e-05, + "loss": 0.0607, + "step": 6560 + }, + { + "epoch": 0.20279029569726525, + "grad_norm": 0.3410740792751312, + "learning_rate": 9.903576630038906e-05, + "loss": 0.0651, + "step": 6570 + }, + { + "epoch": 0.2030989567257238, + "grad_norm": 0.3692129850387573, + "learning_rate": 9.903037290419309e-05, + "loss": 0.0568, + "step": 6580 + }, + { + "epoch": 0.20340761775418237, + "grad_norm": 0.25234097242355347, + "learning_rate": 9.902496461386439e-05, + "loss": 0.0613, + "step": 6590 + }, + { + "epoch": 0.2037162787826409, + "grad_norm": 0.33269834518432617, + "learning_rate": 9.901954143104588e-05, + "loss": 0.0675, + "step": 6600 + }, + { + "epoch": 0.20402493981109945, + "grad_norm": 0.354745477437973, + "learning_rate": 9.901410335738496e-05, + "loss": 0.0631, + "step": 6610 + }, + { + "epoch": 0.204333600839558, + "grad_norm": 0.24259164929389954, + "learning_rate": 9.900865039453358e-05, + "loss": 0.066, + "step": 6620 + }, + { + "epoch": 0.20464226186801654, + "grad_norm": 0.2678314447402954, + "learning_rate": 9.900318254414821e-05, + "loss": 0.0661, + "step": 6630 + }, + { + "epoch": 0.2049509228964751, + "grad_norm": 0.27136746048927307, + "learning_rate": 9.899769980788985e-05, + "loss": 0.0631, + "step": 6640 + }, + { + "epoch": 0.20525958392493363, + "grad_norm": 0.3595856726169586, + "learning_rate": 9.899220218742398e-05, + "loss": 0.0668, + "step": 6650 + }, + { + "epoch": 0.20556824495339218, + "grad_norm": 0.29038500785827637, + "learning_rate": 9.898668968442066e-05, + "loss": 0.0598, + "step": 6660 + }, + { + "epoch": 0.20587690598185074, + "grad_norm": 0.2689511179924011, + "learning_rate": 9.898116230055443e-05, + "loss": 0.0605, + "step": 6670 + }, + { + "epoch": 0.20618556701030927, + "grad_norm": 0.32795020937919617, + "learning_rate": 9.897562003750437e-05, + "loss": 0.0584, + "step": 6680 + }, + { + "epoch": 0.20649422803876782, + "grad_norm": 0.4421314597129822, + "learning_rate": 9.897006289695407e-05, + "loss": 0.061, + "step": 6690 + }, + { + "epoch": 0.20680288906722638, + "grad_norm": 0.4202762246131897, + "learning_rate": 9.896449088059164e-05, + "loss": 0.0586, + "step": 6700 + }, + { + "epoch": 0.2071115500956849, + "grad_norm": 0.3447592258453369, + "learning_rate": 9.89589039901097e-05, + "loss": 0.064, + "step": 6710 + }, + { + "epoch": 0.20742021112414347, + "grad_norm": 0.4347879886627197, + "learning_rate": 9.895330222720542e-05, + "loss": 0.0639, + "step": 6720 + }, + { + "epoch": 0.20772887215260202, + "grad_norm": 0.3698975145816803, + "learning_rate": 9.894768559358047e-05, + "loss": 0.0649, + "step": 6730 + }, + { + "epoch": 0.20803753318106055, + "grad_norm": 0.2494005411863327, + "learning_rate": 9.894205409094101e-05, + "loss": 0.0594, + "step": 6740 + }, + { + "epoch": 0.2083461942095191, + "grad_norm": 0.38694971799850464, + "learning_rate": 9.893640772099777e-05, + "loss": 0.0685, + "step": 6750 + }, + { + "epoch": 0.20865485523797767, + "grad_norm": 0.24023006856441498, + "learning_rate": 9.893074648546595e-05, + "loss": 0.0629, + "step": 6760 + }, + { + "epoch": 0.2089635162664362, + "grad_norm": 0.43136388063430786, + "learning_rate": 9.892507038606528e-05, + "loss": 0.0596, + "step": 6770 + }, + { + "epoch": 0.20927217729489475, + "grad_norm": 0.302676260471344, + "learning_rate": 9.891937942452003e-05, + "loss": 0.0633, + "step": 6780 + }, + { + "epoch": 0.20958083832335328, + "grad_norm": 0.23821811378002167, + "learning_rate": 9.891367360255895e-05, + "loss": 0.0581, + "step": 6790 + }, + { + "epoch": 0.20988949935181184, + "grad_norm": 0.7245838642120361, + "learning_rate": 9.890795292191532e-05, + "loss": 0.0604, + "step": 6800 + }, + { + "epoch": 0.2101981603802704, + "grad_norm": 0.24965906143188477, + "learning_rate": 9.890221738432694e-05, + "loss": 0.0612, + "step": 6810 + }, + { + "epoch": 0.21050682140872892, + "grad_norm": 0.30991825461387634, + "learning_rate": 9.88964669915361e-05, + "loss": 0.0642, + "step": 6820 + }, + { + "epoch": 0.21081548243718748, + "grad_norm": 0.32842084765434265, + "learning_rate": 9.889070174528963e-05, + "loss": 0.0636, + "step": 6830 + }, + { + "epoch": 0.21112414346564604, + "grad_norm": 0.4405710995197296, + "learning_rate": 9.888492164733883e-05, + "loss": 0.0599, + "step": 6840 + }, + { + "epoch": 0.21143280449410456, + "grad_norm": 0.44806480407714844, + "learning_rate": 9.88791266994396e-05, + "loss": 0.0605, + "step": 6850 + }, + { + "epoch": 0.21174146552256312, + "grad_norm": 0.3899080455303192, + "learning_rate": 9.887331690335223e-05, + "loss": 0.0591, + "step": 6860 + }, + { + "epoch": 0.21205012655102168, + "grad_norm": 0.35311540961265564, + "learning_rate": 9.886749226084163e-05, + "loss": 0.0642, + "step": 6870 + }, + { + "epoch": 0.2123587875794802, + "grad_norm": 0.3755998909473419, + "learning_rate": 9.886165277367714e-05, + "loss": 0.0645, + "step": 6880 + }, + { + "epoch": 0.21266744860793876, + "grad_norm": 0.3723001480102539, + "learning_rate": 9.885579844363265e-05, + "loss": 0.0616, + "step": 6890 + }, + { + "epoch": 0.21297610963639732, + "grad_norm": 0.5067237615585327, + "learning_rate": 9.884992927248656e-05, + "loss": 0.0603, + "step": 6900 + }, + { + "epoch": 0.21328477066485585, + "grad_norm": 0.4131212830543518, + "learning_rate": 9.884404526202178e-05, + "loss": 0.0645, + "step": 6910 + }, + { + "epoch": 0.2135934316933144, + "grad_norm": 0.29334500432014465, + "learning_rate": 9.883814641402568e-05, + "loss": 0.06, + "step": 6920 + }, + { + "epoch": 0.21390209272177294, + "grad_norm": 0.3845821022987366, + "learning_rate": 9.88322327302902e-05, + "loss": 0.0633, + "step": 6930 + }, + { + "epoch": 0.2142107537502315, + "grad_norm": 0.42355284094810486, + "learning_rate": 9.882630421261176e-05, + "loss": 0.0582, + "step": 6940 + }, + { + "epoch": 0.21451941477869005, + "grad_norm": 0.27817705273628235, + "learning_rate": 9.88203608627913e-05, + "loss": 0.0641, + "step": 6950 + }, + { + "epoch": 0.21482807580714858, + "grad_norm": 0.379364013671875, + "learning_rate": 9.881440268263422e-05, + "loss": 0.0646, + "step": 6960 + }, + { + "epoch": 0.21513673683560713, + "grad_norm": 0.371061772108078, + "learning_rate": 9.880842967395048e-05, + "loss": 0.0582, + "step": 6970 + }, + { + "epoch": 0.2154453978640657, + "grad_norm": 0.2855817377567291, + "learning_rate": 9.880244183855452e-05, + "loss": 0.0642, + "step": 6980 + }, + { + "epoch": 0.21575405889252422, + "grad_norm": 0.3059510588645935, + "learning_rate": 9.879643917826527e-05, + "loss": 0.0596, + "step": 6990 + }, + { + "epoch": 0.21606271992098278, + "grad_norm": 0.3613103926181793, + "learning_rate": 9.87904216949062e-05, + "loss": 0.0561, + "step": 7000 + }, + { + "epoch": 0.21637138094944133, + "grad_norm": 0.806647777557373, + "learning_rate": 9.878438939030526e-05, + "loss": 0.0636, + "step": 7010 + }, + { + "epoch": 0.21668004197789986, + "grad_norm": 0.29521292448043823, + "learning_rate": 9.877834226629489e-05, + "loss": 0.0574, + "step": 7020 + }, + { + "epoch": 0.21698870300635842, + "grad_norm": 0.38030076026916504, + "learning_rate": 9.877228032471206e-05, + "loss": 0.0564, + "step": 7030 + }, + { + "epoch": 0.21729736403481698, + "grad_norm": 0.4998815655708313, + "learning_rate": 9.876620356739823e-05, + "loss": 0.0582, + "step": 7040 + }, + { + "epoch": 0.2176060250632755, + "grad_norm": 0.5826565027236938, + "learning_rate": 9.876011199619935e-05, + "loss": 0.0591, + "step": 7050 + }, + { + "epoch": 0.21791468609173406, + "grad_norm": 0.3252328336238861, + "learning_rate": 9.875400561296589e-05, + "loss": 0.0556, + "step": 7060 + }, + { + "epoch": 0.21822334712019262, + "grad_norm": 0.47565072774887085, + "learning_rate": 9.874788441955278e-05, + "loss": 0.0648, + "step": 7070 + }, + { + "epoch": 0.21853200814865115, + "grad_norm": 0.32191765308380127, + "learning_rate": 9.874174841781951e-05, + "loss": 0.0588, + "step": 7080 + }, + { + "epoch": 0.2188406691771097, + "grad_norm": 0.36613744497299194, + "learning_rate": 9.873559760963003e-05, + "loss": 0.0625, + "step": 7090 + }, + { + "epoch": 0.21914933020556823, + "grad_norm": 0.3138182759284973, + "learning_rate": 9.872943199685278e-05, + "loss": 0.063, + "step": 7100 + }, + { + "epoch": 0.2194579912340268, + "grad_norm": 0.327452152967453, + "learning_rate": 9.872325158136071e-05, + "loss": 0.0604, + "step": 7110 + }, + { + "epoch": 0.21976665226248535, + "grad_norm": 0.3616715669631958, + "learning_rate": 9.871705636503128e-05, + "loss": 0.0622, + "step": 7120 + }, + { + "epoch": 0.22007531329094387, + "grad_norm": 0.31918463110923767, + "learning_rate": 9.871084634974641e-05, + "loss": 0.0638, + "step": 7130 + }, + { + "epoch": 0.22038397431940243, + "grad_norm": 0.3178709149360657, + "learning_rate": 9.870462153739257e-05, + "loss": 0.0591, + "step": 7140 + }, + { + "epoch": 0.220692635347861, + "grad_norm": 0.24455705285072327, + "learning_rate": 9.869838192986067e-05, + "loss": 0.0555, + "step": 7150 + }, + { + "epoch": 0.22100129637631952, + "grad_norm": 0.33417782187461853, + "learning_rate": 9.869212752904616e-05, + "loss": 0.0632, + "step": 7160 + }, + { + "epoch": 0.22130995740477807, + "grad_norm": 0.28769397735595703, + "learning_rate": 9.868585833684894e-05, + "loss": 0.0643, + "step": 7170 + }, + { + "epoch": 0.22161861843323663, + "grad_norm": 0.5860671997070312, + "learning_rate": 9.867957435517342e-05, + "loss": 0.061, + "step": 7180 + }, + { + "epoch": 0.22192727946169516, + "grad_norm": 0.30399495363235474, + "learning_rate": 9.867327558592854e-05, + "loss": 0.0627, + "step": 7190 + }, + { + "epoch": 0.22223594049015372, + "grad_norm": 0.26259320974349976, + "learning_rate": 9.866696203102766e-05, + "loss": 0.0594, + "step": 7200 + }, + { + "epoch": 0.22254460151861227, + "grad_norm": 0.2799645960330963, + "learning_rate": 9.86606336923887e-05, + "loss": 0.0596, + "step": 7210 + }, + { + "epoch": 0.2228532625470708, + "grad_norm": 0.40772560238838196, + "learning_rate": 9.865429057193403e-05, + "loss": 0.0576, + "step": 7220 + }, + { + "epoch": 0.22316192357552936, + "grad_norm": 0.3454367518424988, + "learning_rate": 9.864793267159053e-05, + "loss": 0.062, + "step": 7230 + }, + { + "epoch": 0.2234705846039879, + "grad_norm": 0.3521794080734253, + "learning_rate": 9.864155999328957e-05, + "loss": 0.0668, + "step": 7240 + }, + { + "epoch": 0.22377924563244644, + "grad_norm": 0.3231131434440613, + "learning_rate": 9.8635172538967e-05, + "loss": 0.0582, + "step": 7250 + }, + { + "epoch": 0.224087906660905, + "grad_norm": 0.2537112236022949, + "learning_rate": 9.862877031056312e-05, + "loss": 0.0601, + "step": 7260 + }, + { + "epoch": 0.22439656768936353, + "grad_norm": 0.35466888546943665, + "learning_rate": 9.862235331002279e-05, + "loss": 0.0658, + "step": 7270 + }, + { + "epoch": 0.2247052287178221, + "grad_norm": 0.3223938047885895, + "learning_rate": 9.861592153929533e-05, + "loss": 0.0608, + "step": 7280 + }, + { + "epoch": 0.22501388974628064, + "grad_norm": 0.3341163098812103, + "learning_rate": 9.860947500033455e-05, + "loss": 0.0658, + "step": 7290 + }, + { + "epoch": 0.22532255077473917, + "grad_norm": 0.2826312482357025, + "learning_rate": 9.86030136950987e-05, + "loss": 0.0608, + "step": 7300 + }, + { + "epoch": 0.22563121180319773, + "grad_norm": 0.22525732219219208, + "learning_rate": 9.85965376255506e-05, + "loss": 0.0618, + "step": 7310 + }, + { + "epoch": 0.22593987283165629, + "grad_norm": 0.3705180585384369, + "learning_rate": 9.859004679365747e-05, + "loss": 0.0666, + "step": 7320 + }, + { + "epoch": 0.22624853386011481, + "grad_norm": 0.34262606501579285, + "learning_rate": 9.858354120139108e-05, + "loss": 0.0581, + "step": 7330 + }, + { + "epoch": 0.22655719488857337, + "grad_norm": 0.28096452355384827, + "learning_rate": 9.857702085072764e-05, + "loss": 0.0597, + "step": 7340 + }, + { + "epoch": 0.22686585591703193, + "grad_norm": 0.2931847870349884, + "learning_rate": 9.857048574364787e-05, + "loss": 0.0574, + "step": 7350 + }, + { + "epoch": 0.22717451694549046, + "grad_norm": 0.3260852098464966, + "learning_rate": 9.856393588213698e-05, + "loss": 0.0645, + "step": 7360 + }, + { + "epoch": 0.227483177973949, + "grad_norm": 0.34116747975349426, + "learning_rate": 9.855737126818458e-05, + "loss": 0.0665, + "step": 7370 + }, + { + "epoch": 0.22779183900240754, + "grad_norm": 0.31618592143058777, + "learning_rate": 9.855079190378491e-05, + "loss": 0.0616, + "step": 7380 + }, + { + "epoch": 0.2281005000308661, + "grad_norm": 0.3245093524456024, + "learning_rate": 9.854419779093655e-05, + "loss": 0.0677, + "step": 7390 + }, + { + "epoch": 0.22840916105932466, + "grad_norm": 0.3634733557701111, + "learning_rate": 9.853758893164264e-05, + "loss": 0.0594, + "step": 7400 + }, + { + "epoch": 0.22871782208778318, + "grad_norm": 0.29311421513557434, + "learning_rate": 9.853096532791078e-05, + "loss": 0.0617, + "step": 7410 + }, + { + "epoch": 0.22902648311624174, + "grad_norm": 0.27706485986709595, + "learning_rate": 9.852432698175304e-05, + "loss": 0.0591, + "step": 7420 + }, + { + "epoch": 0.2293351441447003, + "grad_norm": 0.2861957252025604, + "learning_rate": 9.851767389518597e-05, + "loss": 0.0568, + "step": 7430 + }, + { + "epoch": 0.22964380517315883, + "grad_norm": 0.29124540090560913, + "learning_rate": 9.85110060702306e-05, + "loss": 0.063, + "step": 7440 + }, + { + "epoch": 0.22995246620161738, + "grad_norm": 0.24688906967639923, + "learning_rate": 9.850432350891245e-05, + "loss": 0.0565, + "step": 7450 + }, + { + "epoch": 0.23026112723007594, + "grad_norm": 0.2864222526550293, + "learning_rate": 9.84976262132615e-05, + "loss": 0.0637, + "step": 7460 + }, + { + "epoch": 0.23056978825853447, + "grad_norm": 0.3014529049396515, + "learning_rate": 9.849091418531222e-05, + "loss": 0.0631, + "step": 7470 + }, + { + "epoch": 0.23087844928699303, + "grad_norm": 0.33117973804473877, + "learning_rate": 9.848418742710353e-05, + "loss": 0.058, + "step": 7480 + }, + { + "epoch": 0.23118711031545158, + "grad_norm": 0.3648061752319336, + "learning_rate": 9.847744594067885e-05, + "loss": 0.0602, + "step": 7490 + }, + { + "epoch": 0.2314957713439101, + "grad_norm": 0.3536057472229004, + "learning_rate": 9.847068972808607e-05, + "loss": 0.0616, + "step": 7500 + }, + { + "epoch": 0.23180443237236867, + "grad_norm": 0.3881740868091583, + "learning_rate": 9.846391879137756e-05, + "loss": 0.058, + "step": 7510 + }, + { + "epoch": 0.23211309340082723, + "grad_norm": 0.3524285852909088, + "learning_rate": 9.845713313261012e-05, + "loss": 0.0584, + "step": 7520 + }, + { + "epoch": 0.23242175442928575, + "grad_norm": 0.35331279039382935, + "learning_rate": 9.845033275384505e-05, + "loss": 0.0582, + "step": 7530 + }, + { + "epoch": 0.2327304154577443, + "grad_norm": 0.34950941801071167, + "learning_rate": 9.844351765714818e-05, + "loss": 0.063, + "step": 7540 + }, + { + "epoch": 0.23303907648620284, + "grad_norm": 0.33413413166999817, + "learning_rate": 9.843668784458971e-05, + "loss": 0.0612, + "step": 7550 + }, + { + "epoch": 0.2333477375146614, + "grad_norm": 0.2429344803094864, + "learning_rate": 9.842984331824437e-05, + "loss": 0.0636, + "step": 7560 + }, + { + "epoch": 0.23365639854311995, + "grad_norm": 0.366382360458374, + "learning_rate": 9.842298408019133e-05, + "loss": 0.0592, + "step": 7570 + }, + { + "epoch": 0.23396505957157848, + "grad_norm": 0.2643805742263794, + "learning_rate": 9.841611013251429e-05, + "loss": 0.0578, + "step": 7580 + }, + { + "epoch": 0.23427372060003704, + "grad_norm": 0.3699539005756378, + "learning_rate": 9.840922147730133e-05, + "loss": 0.0601, + "step": 7590 + }, + { + "epoch": 0.2345823816284956, + "grad_norm": 0.7365028858184814, + "learning_rate": 9.840231811664506e-05, + "loss": 0.0606, + "step": 7600 + }, + { + "epoch": 0.23489104265695412, + "grad_norm": 0.37009477615356445, + "learning_rate": 9.839540005264252e-05, + "loss": 0.0578, + "step": 7610 + }, + { + "epoch": 0.23519970368541268, + "grad_norm": 0.3139272630214691, + "learning_rate": 9.838846728739527e-05, + "loss": 0.0583, + "step": 7620 + }, + { + "epoch": 0.23550836471387124, + "grad_norm": 0.37679994106292725, + "learning_rate": 9.838151982300927e-05, + "loss": 0.0614, + "step": 7630 + }, + { + "epoch": 0.23581702574232977, + "grad_norm": 0.39640218019485474, + "learning_rate": 9.8374557661595e-05, + "loss": 0.0634, + "step": 7640 + }, + { + "epoch": 0.23612568677078832, + "grad_norm": 0.2995794415473938, + "learning_rate": 9.836758080526735e-05, + "loss": 0.06, + "step": 7650 + }, + { + "epoch": 0.23643434779924688, + "grad_norm": 0.3081921935081482, + "learning_rate": 9.836058925614575e-05, + "loss": 0.068, + "step": 7660 + }, + { + "epoch": 0.2367430088277054, + "grad_norm": 0.3766205608844757, + "learning_rate": 9.8353583016354e-05, + "loss": 0.0685, + "step": 7670 + }, + { + "epoch": 0.23705166985616397, + "grad_norm": 0.3426515758037567, + "learning_rate": 9.834656208802044e-05, + "loss": 0.0567, + "step": 7680 + }, + { + "epoch": 0.2373603308846225, + "grad_norm": 0.3037927746772766, + "learning_rate": 9.833952647327784e-05, + "loss": 0.057, + "step": 7690 + }, + { + "epoch": 0.23766899191308105, + "grad_norm": 0.2608187794685364, + "learning_rate": 9.833247617426342e-05, + "loss": 0.0645, + "step": 7700 + }, + { + "epoch": 0.2379776529415396, + "grad_norm": 0.2682100236415863, + "learning_rate": 9.832541119311889e-05, + "loss": 0.0609, + "step": 7710 + }, + { + "epoch": 0.23828631396999814, + "grad_norm": 0.2914521098136902, + "learning_rate": 9.83183315319904e-05, + "loss": 0.0573, + "step": 7720 + }, + { + "epoch": 0.2385949749984567, + "grad_norm": 0.40061914920806885, + "learning_rate": 9.831123719302855e-05, + "loss": 0.0602, + "step": 7730 + }, + { + "epoch": 0.23890363602691525, + "grad_norm": 0.3374808132648468, + "learning_rate": 9.830412817838842e-05, + "loss": 0.0566, + "step": 7740 + }, + { + "epoch": 0.23921229705537378, + "grad_norm": 0.30275481939315796, + "learning_rate": 9.829700449022956e-05, + "loss": 0.0606, + "step": 7750 + }, + { + "epoch": 0.23952095808383234, + "grad_norm": 0.37703490257263184, + "learning_rate": 9.828986613071593e-05, + "loss": 0.0572, + "step": 7760 + }, + { + "epoch": 0.2398296191122909, + "grad_norm": 0.290799617767334, + "learning_rate": 9.828271310201601e-05, + "loss": 0.0649, + "step": 7770 + }, + { + "epoch": 0.24013828014074942, + "grad_norm": 0.29602065682411194, + "learning_rate": 9.827554540630268e-05, + "loss": 0.058, + "step": 7780 + }, + { + "epoch": 0.24044694116920798, + "grad_norm": 0.3123007118701935, + "learning_rate": 9.826836304575329e-05, + "loss": 0.0593, + "step": 7790 + }, + { + "epoch": 0.24075560219766654, + "grad_norm": 0.3341997265815735, + "learning_rate": 9.826116602254966e-05, + "loss": 0.0574, + "step": 7800 + }, + { + "epoch": 0.24106426322612506, + "grad_norm": 0.424597829580307, + "learning_rate": 9.825395433887805e-05, + "loss": 0.0596, + "step": 7810 + }, + { + "epoch": 0.24137292425458362, + "grad_norm": 0.3122120797634125, + "learning_rate": 9.824672799692917e-05, + "loss": 0.0566, + "step": 7820 + }, + { + "epoch": 0.24168158528304215, + "grad_norm": 0.35231199860572815, + "learning_rate": 9.823948699889823e-05, + "loss": 0.0618, + "step": 7830 + }, + { + "epoch": 0.2419902463115007, + "grad_norm": 0.3178722560405731, + "learning_rate": 9.823223134698483e-05, + "loss": 0.0594, + "step": 7840 + }, + { + "epoch": 0.24229890733995926, + "grad_norm": 0.3863544464111328, + "learning_rate": 9.822496104339303e-05, + "loss": 0.0638, + "step": 7850 + }, + { + "epoch": 0.2426075683684178, + "grad_norm": 0.3208511769771576, + "learning_rate": 9.821767609033138e-05, + "loss": 0.059, + "step": 7860 + }, + { + "epoch": 0.24291622939687635, + "grad_norm": 0.37483277916908264, + "learning_rate": 9.821037649001284e-05, + "loss": 0.058, + "step": 7870 + }, + { + "epoch": 0.2432248904253349, + "grad_norm": 0.37077072262763977, + "learning_rate": 9.820306224465486e-05, + "loss": 0.0615, + "step": 7880 + }, + { + "epoch": 0.24353355145379343, + "grad_norm": 0.35546112060546875, + "learning_rate": 9.819573335647928e-05, + "loss": 0.0625, + "step": 7890 + }, + { + "epoch": 0.243842212482252, + "grad_norm": 0.43891632556915283, + "learning_rate": 9.818838982771246e-05, + "loss": 0.0635, + "step": 7900 + }, + { + "epoch": 0.24415087351071055, + "grad_norm": 0.34409773349761963, + "learning_rate": 9.818103166058514e-05, + "loss": 0.0585, + "step": 7910 + }, + { + "epoch": 0.24445953453916908, + "grad_norm": 0.30644163489341736, + "learning_rate": 9.817365885733254e-05, + "loss": 0.062, + "step": 7920 + }, + { + "epoch": 0.24476819556762763, + "grad_norm": 0.4399518370628357, + "learning_rate": 9.816627142019434e-05, + "loss": 0.0613, + "step": 7930 + }, + { + "epoch": 0.2450768565960862, + "grad_norm": 0.30505403876304626, + "learning_rate": 9.815886935141463e-05, + "loss": 0.0623, + "step": 7940 + }, + { + "epoch": 0.24538551762454472, + "grad_norm": 0.3379201292991638, + "learning_rate": 9.8151452653242e-05, + "loss": 0.0586, + "step": 7950 + }, + { + "epoch": 0.24569417865300328, + "grad_norm": 0.6844927072525024, + "learning_rate": 9.814402132792939e-05, + "loss": 0.0639, + "step": 7960 + }, + { + "epoch": 0.24600283968146183, + "grad_norm": 0.3249509930610657, + "learning_rate": 9.813657537773428e-05, + "loss": 0.0631, + "step": 7970 + }, + { + "epoch": 0.24631150070992036, + "grad_norm": 0.3393564224243164, + "learning_rate": 9.812911480491854e-05, + "loss": 0.0616, + "step": 7980 + }, + { + "epoch": 0.24662016173837892, + "grad_norm": 0.3404988646507263, + "learning_rate": 9.81216396117485e-05, + "loss": 0.0573, + "step": 7990 + }, + { + "epoch": 0.24692882276683745, + "grad_norm": 0.33501455187797546, + "learning_rate": 9.811414980049491e-05, + "loss": 0.0569, + "step": 8000 + }, + { + "epoch": 0.247237483795296, + "grad_norm": 0.2497110366821289, + "learning_rate": 9.810664537343301e-05, + "loss": 0.0527, + "step": 8010 + }, + { + "epoch": 0.24754614482375456, + "grad_norm": 0.30973368883132935, + "learning_rate": 9.809912633284243e-05, + "loss": 0.0598, + "step": 8020 + }, + { + "epoch": 0.2478548058522131, + "grad_norm": 0.3260195851325989, + "learning_rate": 9.809159268100725e-05, + "loss": 0.0601, + "step": 8030 + }, + { + "epoch": 0.24816346688067165, + "grad_norm": 0.24536465108394623, + "learning_rate": 9.808404442021599e-05, + "loss": 0.0627, + "step": 8040 + }, + { + "epoch": 0.2484721279091302, + "grad_norm": 0.26544252038002014, + "learning_rate": 9.807648155276163e-05, + "loss": 0.0582, + "step": 8050 + }, + { + "epoch": 0.24878078893758873, + "grad_norm": 0.48226019740104675, + "learning_rate": 9.806890408094156e-05, + "loss": 0.0614, + "step": 8060 + }, + { + "epoch": 0.2490894499660473, + "grad_norm": 0.2965430021286011, + "learning_rate": 9.806131200705761e-05, + "loss": 0.0583, + "step": 8070 + }, + { + "epoch": 0.24939811099450584, + "grad_norm": 0.3296026885509491, + "learning_rate": 9.805370533341605e-05, + "loss": 0.0576, + "step": 8080 + }, + { + "epoch": 0.24970677202296437, + "grad_norm": 0.44993799924850464, + "learning_rate": 9.804608406232762e-05, + "loss": 0.06, + "step": 8090 + }, + { + "epoch": 0.25001543305142293, + "grad_norm": 0.3844868838787079, + "learning_rate": 9.803844819610741e-05, + "loss": 0.0591, + "step": 8100 + }, + { + "epoch": 0.2503240940798815, + "grad_norm": 0.2610471844673157, + "learning_rate": 9.803079773707504e-05, + "loss": 0.057, + "step": 8110 + }, + { + "epoch": 0.25063275510834004, + "grad_norm": 0.2830440104007721, + "learning_rate": 9.802313268755447e-05, + "loss": 0.0614, + "step": 8120 + }, + { + "epoch": 0.25094141613679855, + "grad_norm": 0.3512442708015442, + "learning_rate": 9.801545304987419e-05, + "loss": 0.0595, + "step": 8130 + }, + { + "epoch": 0.2512500771652571, + "grad_norm": 0.27959370613098145, + "learning_rate": 9.800775882636704e-05, + "loss": 0.0607, + "step": 8140 + }, + { + "epoch": 0.25155873819371566, + "grad_norm": 0.40173211693763733, + "learning_rate": 9.800005001937034e-05, + "loss": 0.06, + "step": 8150 + }, + { + "epoch": 0.2518673992221742, + "grad_norm": 0.3110259473323822, + "learning_rate": 9.79923266312258e-05, + "loss": 0.0606, + "step": 8160 + }, + { + "epoch": 0.25217606025063277, + "grad_norm": 0.33174124360084534, + "learning_rate": 9.79845886642796e-05, + "loss": 0.0621, + "step": 8170 + }, + { + "epoch": 0.25248472127909133, + "grad_norm": 0.2962331473827362, + "learning_rate": 9.797683612088233e-05, + "loss": 0.0578, + "step": 8180 + }, + { + "epoch": 0.25279338230754983, + "grad_norm": 0.332874059677124, + "learning_rate": 9.796906900338898e-05, + "loss": 0.0594, + "step": 8190 + }, + { + "epoch": 0.2531020433360084, + "grad_norm": 0.3059181869029999, + "learning_rate": 9.796128731415903e-05, + "loss": 0.0583, + "step": 8200 + }, + { + "epoch": 0.25341070436446694, + "grad_norm": 0.3392615020275116, + "learning_rate": 9.795349105555634e-05, + "loss": 0.0612, + "step": 8210 + }, + { + "epoch": 0.2537193653929255, + "grad_norm": 0.3160492479801178, + "learning_rate": 9.794568022994922e-05, + "loss": 0.0559, + "step": 8220 + }, + { + "epoch": 0.25402802642138406, + "grad_norm": 0.6149255633354187, + "learning_rate": 9.793785483971034e-05, + "loss": 0.058, + "step": 8230 + }, + { + "epoch": 0.25433668744984256, + "grad_norm": 0.2725014090538025, + "learning_rate": 9.793001488721691e-05, + "loss": 0.0604, + "step": 8240 + }, + { + "epoch": 0.2546453484783011, + "grad_norm": 0.26116836071014404, + "learning_rate": 9.792216037485047e-05, + "loss": 0.0598, + "step": 8250 + }, + { + "epoch": 0.25495400950675967, + "grad_norm": 0.34398332238197327, + "learning_rate": 9.791429130499704e-05, + "loss": 0.0581, + "step": 8260 + }, + { + "epoch": 0.25526267053521823, + "grad_norm": 0.3979973793029785, + "learning_rate": 9.790640768004698e-05, + "loss": 0.0576, + "step": 8270 + }, + { + "epoch": 0.2555713315636768, + "grad_norm": 0.4290270209312439, + "learning_rate": 9.789850950239518e-05, + "loss": 0.0586, + "step": 8280 + }, + { + "epoch": 0.25587999259213534, + "grad_norm": 0.2969729006290436, + "learning_rate": 9.789059677444089e-05, + "loss": 0.0652, + "step": 8290 + }, + { + "epoch": 0.25618865362059384, + "grad_norm": 0.3432493507862091, + "learning_rate": 9.788266949858776e-05, + "loss": 0.0602, + "step": 8300 + }, + { + "epoch": 0.2564973146490524, + "grad_norm": 0.34590160846710205, + "learning_rate": 9.787472767724392e-05, + "loss": 0.0606, + "step": 8310 + }, + { + "epoch": 0.25680597567751096, + "grad_norm": 0.35182759165763855, + "learning_rate": 9.786677131282185e-05, + "loss": 0.0587, + "step": 8320 + }, + { + "epoch": 0.2571146367059695, + "grad_norm": 0.3782471716403961, + "learning_rate": 9.785880040773853e-05, + "loss": 0.0591, + "step": 8330 + }, + { + "epoch": 0.25742329773442807, + "grad_norm": 0.38115954399108887, + "learning_rate": 9.785081496441527e-05, + "loss": 0.0586, + "step": 8340 + }, + { + "epoch": 0.25773195876288657, + "grad_norm": 0.3423364758491516, + "learning_rate": 9.784281498527785e-05, + "loss": 0.0607, + "step": 8350 + }, + { + "epoch": 0.2580406197913451, + "grad_norm": 0.3824802041053772, + "learning_rate": 9.783480047275646e-05, + "loss": 0.0564, + "step": 8360 + }, + { + "epoch": 0.2583492808198037, + "grad_norm": 0.35345107316970825, + "learning_rate": 9.78267714292857e-05, + "loss": 0.0513, + "step": 8370 + }, + { + "epoch": 0.25865794184826224, + "grad_norm": 1.2818603515625, + "learning_rate": 9.781872785730454e-05, + "loss": 0.0576, + "step": 8380 + }, + { + "epoch": 0.2589666028767208, + "grad_norm": 0.3998434841632843, + "learning_rate": 9.781066975925646e-05, + "loss": 0.0632, + "step": 8390 + }, + { + "epoch": 0.25927526390517935, + "grad_norm": 0.22661836445331573, + "learning_rate": 9.780259713758928e-05, + "loss": 0.0586, + "step": 8400 + }, + { + "epoch": 0.25958392493363786, + "grad_norm": 0.2250576913356781, + "learning_rate": 9.779450999475524e-05, + "loss": 0.0599, + "step": 8410 + }, + { + "epoch": 0.2598925859620964, + "grad_norm": 0.3395307958126068, + "learning_rate": 9.7786408333211e-05, + "loss": 0.0617, + "step": 8420 + }, + { + "epoch": 0.26020124699055497, + "grad_norm": 0.336107075214386, + "learning_rate": 9.777829215541764e-05, + "loss": 0.0566, + "step": 8430 + }, + { + "epoch": 0.2605099080190135, + "grad_norm": 0.3653830587863922, + "learning_rate": 9.777016146384064e-05, + "loss": 0.0625, + "step": 8440 + }, + { + "epoch": 0.2608185690474721, + "grad_norm": 0.2860945165157318, + "learning_rate": 9.776201626094988e-05, + "loss": 0.0686, + "step": 8450 + }, + { + "epoch": 0.26112723007593064, + "grad_norm": 0.32020577788352966, + "learning_rate": 9.775385654921965e-05, + "loss": 0.0588, + "step": 8460 + }, + { + "epoch": 0.26143589110438914, + "grad_norm": 0.3276720643043518, + "learning_rate": 9.774568233112868e-05, + "loss": 0.0597, + "step": 8470 + }, + { + "epoch": 0.2617445521328477, + "grad_norm": 0.34632813930511475, + "learning_rate": 9.773749360916007e-05, + "loss": 0.0607, + "step": 8480 + }, + { + "epoch": 0.26205321316130625, + "grad_norm": 0.36245718598365784, + "learning_rate": 9.772929038580134e-05, + "loss": 0.063, + "step": 8490 + }, + { + "epoch": 0.2623618741897648, + "grad_norm": 0.2944202423095703, + "learning_rate": 9.772107266354439e-05, + "loss": 0.0612, + "step": 8500 + }, + { + "epoch": 0.26267053521822337, + "grad_norm": 0.3674871027469635, + "learning_rate": 9.77128404448856e-05, + "loss": 0.0601, + "step": 8510 + }, + { + "epoch": 0.26297919624668187, + "grad_norm": 0.2862493693828583, + "learning_rate": 9.770459373232565e-05, + "loss": 0.0595, + "step": 8520 + }, + { + "epoch": 0.2632878572751404, + "grad_norm": 0.3383352756500244, + "learning_rate": 9.769633252836969e-05, + "loss": 0.0551, + "step": 8530 + }, + { + "epoch": 0.263596518303599, + "grad_norm": 0.42922040820121765, + "learning_rate": 9.768805683552724e-05, + "loss": 0.0619, + "step": 8540 + }, + { + "epoch": 0.26390517933205754, + "grad_norm": 0.5190513134002686, + "learning_rate": 9.767976665631228e-05, + "loss": 0.055, + "step": 8550 + }, + { + "epoch": 0.2642138403605161, + "grad_norm": 0.395084410905838, + "learning_rate": 9.767146199324311e-05, + "loss": 0.0618, + "step": 8560 + }, + { + "epoch": 0.26452250138897465, + "grad_norm": 0.29816243052482605, + "learning_rate": 9.766314284884249e-05, + "loss": 0.0641, + "step": 8570 + }, + { + "epoch": 0.26483116241743315, + "grad_norm": 0.3626165986061096, + "learning_rate": 9.765480922563752e-05, + "loss": 0.0597, + "step": 8580 + }, + { + "epoch": 0.2651398234458917, + "grad_norm": 0.33852964639663696, + "learning_rate": 9.764646112615978e-05, + "loss": 0.0603, + "step": 8590 + }, + { + "epoch": 0.26544848447435027, + "grad_norm": 0.3976552486419678, + "learning_rate": 9.763809855294517e-05, + "loss": 0.0575, + "step": 8600 + }, + { + "epoch": 0.2657571455028088, + "grad_norm": 0.4344286024570465, + "learning_rate": 9.762972150853404e-05, + "loss": 0.0605, + "step": 8610 + }, + { + "epoch": 0.2660658065312674, + "grad_norm": 0.29477259516716003, + "learning_rate": 9.762132999547111e-05, + "loss": 0.0572, + "step": 8620 + }, + { + "epoch": 0.26637446755972594, + "grad_norm": 0.4503686726093292, + "learning_rate": 9.761292401630549e-05, + "loss": 0.061, + "step": 8630 + }, + { + "epoch": 0.26668312858818444, + "grad_norm": 0.3200767934322357, + "learning_rate": 9.76045035735907e-05, + "loss": 0.0633, + "step": 8640 + }, + { + "epoch": 0.266991789616643, + "grad_norm": 0.27824628353118896, + "learning_rate": 9.759606866988464e-05, + "loss": 0.0632, + "step": 8650 + }, + { + "epoch": 0.26730045064510155, + "grad_norm": 0.29770058393478394, + "learning_rate": 9.758761930774963e-05, + "loss": 0.0567, + "step": 8660 + }, + { + "epoch": 0.2676091116735601, + "grad_norm": 0.2907274067401886, + "learning_rate": 9.757915548975235e-05, + "loss": 0.0629, + "step": 8670 + }, + { + "epoch": 0.26791777270201866, + "grad_norm": 0.31242290139198303, + "learning_rate": 9.757067721846389e-05, + "loss": 0.0626, + "step": 8680 + }, + { + "epoch": 0.26822643373047717, + "grad_norm": 0.3045295476913452, + "learning_rate": 9.756218449645971e-05, + "loss": 0.0543, + "step": 8690 + }, + { + "epoch": 0.2685350947589357, + "grad_norm": 0.34361398220062256, + "learning_rate": 9.75536773263197e-05, + "loss": 0.0557, + "step": 8700 + }, + { + "epoch": 0.2688437557873943, + "grad_norm": 0.280612051486969, + "learning_rate": 9.75451557106281e-05, + "loss": 0.0627, + "step": 8710 + }, + { + "epoch": 0.26915241681585284, + "grad_norm": 0.32431480288505554, + "learning_rate": 9.753661965197354e-05, + "loss": 0.0595, + "step": 8720 + }, + { + "epoch": 0.2694610778443114, + "grad_norm": 0.21596777439117432, + "learning_rate": 9.752806915294908e-05, + "loss": 0.059, + "step": 8730 + }, + { + "epoch": 0.26976973887276995, + "grad_norm": 0.25313621759414673, + "learning_rate": 9.75195042161521e-05, + "loss": 0.0582, + "step": 8740 + }, + { + "epoch": 0.27007839990122845, + "grad_norm": 0.3264451324939728, + "learning_rate": 9.751092484418442e-05, + "loss": 0.0547, + "step": 8750 + }, + { + "epoch": 0.270387060929687, + "grad_norm": 0.30401062965393066, + "learning_rate": 9.750233103965224e-05, + "loss": 0.0571, + "step": 8760 + }, + { + "epoch": 0.27069572195814556, + "grad_norm": 0.33266761898994446, + "learning_rate": 9.749372280516611e-05, + "loss": 0.0641, + "step": 8770 + }, + { + "epoch": 0.2710043829866041, + "grad_norm": 0.3364381492137909, + "learning_rate": 9.748510014334097e-05, + "loss": 0.0595, + "step": 8780 + }, + { + "epoch": 0.2713130440150627, + "grad_norm": 0.30498018860816956, + "learning_rate": 9.747646305679621e-05, + "loss": 0.0614, + "step": 8790 + }, + { + "epoch": 0.2716217050435212, + "grad_norm": 0.375044584274292, + "learning_rate": 9.74678115481555e-05, + "loss": 0.053, + "step": 8800 + }, + { + "epoch": 0.27193036607197973, + "grad_norm": 0.4384313225746155, + "learning_rate": 9.745914562004696e-05, + "loss": 0.0545, + "step": 8810 + }, + { + "epoch": 0.2722390271004383, + "grad_norm": 0.29880690574645996, + "learning_rate": 9.745046527510307e-05, + "loss": 0.0575, + "step": 8820 + }, + { + "epoch": 0.27254768812889685, + "grad_norm": 0.3265838027000427, + "learning_rate": 9.744177051596068e-05, + "loss": 0.0584, + "step": 8830 + }, + { + "epoch": 0.2728563491573554, + "grad_norm": 0.4497830867767334, + "learning_rate": 9.743306134526105e-05, + "loss": 0.0606, + "step": 8840 + }, + { + "epoch": 0.27316501018581396, + "grad_norm": 0.32285311818122864, + "learning_rate": 9.742433776564977e-05, + "loss": 0.0623, + "step": 8850 + }, + { + "epoch": 0.27347367121427246, + "grad_norm": 0.3437435030937195, + "learning_rate": 9.741559977977683e-05, + "loss": 0.0569, + "step": 8860 + }, + { + "epoch": 0.273782332242731, + "grad_norm": 0.3379465341567993, + "learning_rate": 9.740684739029661e-05, + "loss": 0.0623, + "step": 8870 + }, + { + "epoch": 0.2740909932711896, + "grad_norm": 0.431861013174057, + "learning_rate": 9.739808059986789e-05, + "loss": 0.0552, + "step": 8880 + }, + { + "epoch": 0.27439965429964813, + "grad_norm": 0.4069928228855133, + "learning_rate": 9.738929941115373e-05, + "loss": 0.0676, + "step": 8890 + }, + { + "epoch": 0.2747083153281067, + "grad_norm": 1.2332415580749512, + "learning_rate": 9.738050382682167e-05, + "loss": 0.0592, + "step": 8900 + }, + { + "epoch": 0.27501697635656525, + "grad_norm": 0.4783972501754761, + "learning_rate": 9.737169384954355e-05, + "loss": 0.0568, + "step": 8910 + }, + { + "epoch": 0.27532563738502375, + "grad_norm": 0.398701548576355, + "learning_rate": 9.736286948199562e-05, + "loss": 0.0587, + "step": 8920 + }, + { + "epoch": 0.2756342984134823, + "grad_norm": 0.48909980058670044, + "learning_rate": 9.735403072685848e-05, + "loss": 0.0648, + "step": 8930 + }, + { + "epoch": 0.27594295944194086, + "grad_norm": 0.34863007068634033, + "learning_rate": 9.734517758681712e-05, + "loss": 0.0597, + "step": 8940 + }, + { + "epoch": 0.2762516204703994, + "grad_norm": 0.25364580750465393, + "learning_rate": 9.733631006456088e-05, + "loss": 0.0595, + "step": 8950 + }, + { + "epoch": 0.276560281498858, + "grad_norm": 0.3115352690219879, + "learning_rate": 9.732742816278348e-05, + "loss": 0.0605, + "step": 8960 + }, + { + "epoch": 0.2768689425273165, + "grad_norm": 0.36826545000076294, + "learning_rate": 9.731853188418302e-05, + "loss": 0.0585, + "step": 8970 + }, + { + "epoch": 0.27717760355577503, + "grad_norm": 0.46914899349212646, + "learning_rate": 9.730962123146194e-05, + "loss": 0.0615, + "step": 8980 + }, + { + "epoch": 0.2774862645842336, + "grad_norm": 0.2879350781440735, + "learning_rate": 9.730069620732709e-05, + "loss": 0.0613, + "step": 8990 + }, + { + "epoch": 0.27779492561269215, + "grad_norm": 0.36344975233078003, + "learning_rate": 9.72917568144896e-05, + "loss": 0.0636, + "step": 9000 + }, + { + "epoch": 0.2781035866411507, + "grad_norm": 0.31350913643836975, + "learning_rate": 9.728280305566509e-05, + "loss": 0.0616, + "step": 9010 + }, + { + "epoch": 0.27841224766960926, + "grad_norm": 0.3369823098182678, + "learning_rate": 9.727383493357343e-05, + "loss": 0.0598, + "step": 9020 + }, + { + "epoch": 0.27872090869806776, + "grad_norm": 0.469749391078949, + "learning_rate": 9.726485245093891e-05, + "loss": 0.0619, + "step": 9030 + }, + { + "epoch": 0.2790295697265263, + "grad_norm": 0.33442333340644836, + "learning_rate": 9.725585561049018e-05, + "loss": 0.0586, + "step": 9040 + }, + { + "epoch": 0.2793382307549849, + "grad_norm": 0.31244105100631714, + "learning_rate": 9.724684441496022e-05, + "loss": 0.059, + "step": 9050 + }, + { + "epoch": 0.27964689178344343, + "grad_norm": 0.25077810883522034, + "learning_rate": 9.72378188670864e-05, + "loss": 0.0674, + "step": 9060 + }, + { + "epoch": 0.279955552811902, + "grad_norm": 0.4921092987060547, + "learning_rate": 9.722877896961047e-05, + "loss": 0.0613, + "step": 9070 + }, + { + "epoch": 0.28026421384036054, + "grad_norm": 0.2832631468772888, + "learning_rate": 9.721972472527848e-05, + "loss": 0.0594, + "step": 9080 + }, + { + "epoch": 0.28057287486881904, + "grad_norm": 0.28690436482429504, + "learning_rate": 9.721065613684089e-05, + "loss": 0.0577, + "step": 9090 + }, + { + "epoch": 0.2808815358972776, + "grad_norm": 0.3569124937057495, + "learning_rate": 9.72015732070525e-05, + "loss": 0.0588, + "step": 9100 + }, + { + "epoch": 0.28119019692573616, + "grad_norm": 0.2699451148509979, + "learning_rate": 9.719247593867244e-05, + "loss": 0.0565, + "step": 9110 + }, + { + "epoch": 0.2814988579541947, + "grad_norm": 0.3332992494106293, + "learning_rate": 9.718336433446423e-05, + "loss": 0.0588, + "step": 9120 + }, + { + "epoch": 0.28180751898265327, + "grad_norm": 0.22008907794952393, + "learning_rate": 9.717423839719574e-05, + "loss": 0.0624, + "step": 9130 + }, + { + "epoch": 0.2821161800111118, + "grad_norm": 0.26886826753616333, + "learning_rate": 9.71650981296392e-05, + "loss": 0.0547, + "step": 9140 + }, + { + "epoch": 0.28242484103957033, + "grad_norm": 0.2991245985031128, + "learning_rate": 9.715594353457118e-05, + "loss": 0.0642, + "step": 9150 + }, + { + "epoch": 0.2827335020680289, + "grad_norm": 0.2826148271560669, + "learning_rate": 9.714677461477257e-05, + "loss": 0.054, + "step": 9160 + }, + { + "epoch": 0.28304216309648744, + "grad_norm": 0.4426681697368622, + "learning_rate": 9.713759137302869e-05, + "loss": 0.0583, + "step": 9170 + }, + { + "epoch": 0.283350824124946, + "grad_norm": 0.2492821216583252, + "learning_rate": 9.712839381212914e-05, + "loss": 0.0535, + "step": 9180 + }, + { + "epoch": 0.28365948515340456, + "grad_norm": 0.38565051555633545, + "learning_rate": 9.71191819348679e-05, + "loss": 0.064, + "step": 9190 + }, + { + "epoch": 0.28396814618186306, + "grad_norm": 0.26531732082366943, + "learning_rate": 9.710995574404331e-05, + "loss": 0.0601, + "step": 9200 + }, + { + "epoch": 0.2842768072103216, + "grad_norm": 0.3432020843029022, + "learning_rate": 9.710071524245802e-05, + "loss": 0.0595, + "step": 9210 + }, + { + "epoch": 0.28458546823878017, + "grad_norm": 0.3680887222290039, + "learning_rate": 9.709146043291906e-05, + "loss": 0.0617, + "step": 9220 + }, + { + "epoch": 0.2848941292672387, + "grad_norm": 0.3011508285999298, + "learning_rate": 9.70821913182378e-05, + "loss": 0.0596, + "step": 9230 + }, + { + "epoch": 0.2852027902956973, + "grad_norm": 0.24060305953025818, + "learning_rate": 9.707290790122995e-05, + "loss": 0.0568, + "step": 9240 + }, + { + "epoch": 0.2855114513241558, + "grad_norm": 0.37524428963661194, + "learning_rate": 9.706361018471557e-05, + "loss": 0.0606, + "step": 9250 + }, + { + "epoch": 0.28582011235261434, + "grad_norm": 0.37336403131484985, + "learning_rate": 9.705429817151906e-05, + "loss": 0.0609, + "step": 9260 + }, + { + "epoch": 0.2861287733810729, + "grad_norm": 0.31784290075302124, + "learning_rate": 9.704497186446917e-05, + "loss": 0.0613, + "step": 9270 + }, + { + "epoch": 0.28643743440953146, + "grad_norm": 0.45595690608024597, + "learning_rate": 9.703563126639896e-05, + "loss": 0.0601, + "step": 9280 + }, + { + "epoch": 0.28674609543799, + "grad_norm": 0.4099086821079254, + "learning_rate": 9.70262763801459e-05, + "loss": 0.0564, + "step": 9290 + }, + { + "epoch": 0.28705475646644857, + "grad_norm": 0.32722899317741394, + "learning_rate": 9.701690720855171e-05, + "loss": 0.0622, + "step": 9300 + }, + { + "epoch": 0.28736341749490707, + "grad_norm": 0.3858245313167572, + "learning_rate": 9.700752375446253e-05, + "loss": 0.0567, + "step": 9310 + }, + { + "epoch": 0.2876720785233656, + "grad_norm": 0.37717944383621216, + "learning_rate": 9.69981260207288e-05, + "loss": 0.0534, + "step": 9320 + }, + { + "epoch": 0.2879807395518242, + "grad_norm": 0.428760290145874, + "learning_rate": 9.698871401020529e-05, + "loss": 0.0559, + "step": 9330 + }, + { + "epoch": 0.28828940058028274, + "grad_norm": 0.31394684314727783, + "learning_rate": 9.697928772575112e-05, + "loss": 0.0617, + "step": 9340 + }, + { + "epoch": 0.2885980616087413, + "grad_norm": 0.33836838603019714, + "learning_rate": 9.696984717022976e-05, + "loss": 0.06, + "step": 9350 + }, + { + "epoch": 0.28890672263719985, + "grad_norm": 0.5426814556121826, + "learning_rate": 9.6960392346509e-05, + "loss": 0.0621, + "step": 9360 + }, + { + "epoch": 0.28921538366565835, + "grad_norm": 0.3425922989845276, + "learning_rate": 9.695092325746097e-05, + "loss": 0.0542, + "step": 9370 + }, + { + "epoch": 0.2895240446941169, + "grad_norm": 0.268331378698349, + "learning_rate": 9.694143990596211e-05, + "loss": 0.0555, + "step": 9380 + }, + { + "epoch": 0.28983270572257547, + "grad_norm": 0.3338371217250824, + "learning_rate": 9.693194229489325e-05, + "loss": 0.0602, + "step": 9390 + }, + { + "epoch": 0.290141366751034, + "grad_norm": 0.26813289523124695, + "learning_rate": 9.692243042713944e-05, + "loss": 0.059, + "step": 9400 + }, + { + "epoch": 0.2904500277794926, + "grad_norm": 0.33067402243614197, + "learning_rate": 9.691290430559022e-05, + "loss": 0.0606, + "step": 9410 + }, + { + "epoch": 0.2907586888079511, + "grad_norm": 0.32516223192214966, + "learning_rate": 9.690336393313932e-05, + "loss": 0.0539, + "step": 9420 + }, + { + "epoch": 0.29106734983640964, + "grad_norm": 0.2531454861164093, + "learning_rate": 9.689380931268487e-05, + "loss": 0.0582, + "step": 9430 + }, + { + "epoch": 0.2913760108648682, + "grad_norm": 0.3072642683982849, + "learning_rate": 9.688424044712932e-05, + "loss": 0.0583, + "step": 9440 + }, + { + "epoch": 0.29168467189332675, + "grad_norm": 0.31331902742385864, + "learning_rate": 9.687465733937942e-05, + "loss": 0.0527, + "step": 9450 + }, + { + "epoch": 0.2919933329217853, + "grad_norm": 0.34373629093170166, + "learning_rate": 9.686505999234627e-05, + "loss": 0.0597, + "step": 9460 + }, + { + "epoch": 0.29230199395024387, + "grad_norm": 0.3271065950393677, + "learning_rate": 9.685544840894529e-05, + "loss": 0.0574, + "step": 9470 + }, + { + "epoch": 0.29261065497870237, + "grad_norm": 0.29762959480285645, + "learning_rate": 9.684582259209624e-05, + "loss": 0.0574, + "step": 9480 + }, + { + "epoch": 0.2929193160071609, + "grad_norm": 0.2751370668411255, + "learning_rate": 9.683618254472317e-05, + "loss": 0.0543, + "step": 9490 + }, + { + "epoch": 0.2932279770356195, + "grad_norm": 0.36776623129844666, + "learning_rate": 9.682652826975449e-05, + "loss": 0.0498, + "step": 9500 + }, + { + "epoch": 0.29353663806407804, + "grad_norm": 0.34075260162353516, + "learning_rate": 9.681685977012291e-05, + "loss": 0.0578, + "step": 9510 + }, + { + "epoch": 0.2938452990925366, + "grad_norm": 0.3308494985103607, + "learning_rate": 9.680717704876546e-05, + "loss": 0.062, + "step": 9520 + }, + { + "epoch": 0.29415396012099515, + "grad_norm": 0.40597182512283325, + "learning_rate": 9.679748010862349e-05, + "loss": 0.0531, + "step": 9530 + }, + { + "epoch": 0.29446262114945365, + "grad_norm": 0.33759605884552, + "learning_rate": 9.678776895264267e-05, + "loss": 0.0611, + "step": 9540 + }, + { + "epoch": 0.2947712821779122, + "grad_norm": 0.3181621730327606, + "learning_rate": 9.6778043583773e-05, + "loss": 0.056, + "step": 9550 + }, + { + "epoch": 0.29507994320637077, + "grad_norm": 0.4236881136894226, + "learning_rate": 9.67683040049688e-05, + "loss": 0.0528, + "step": 9560 + }, + { + "epoch": 0.2953886042348293, + "grad_norm": 0.3441140353679657, + "learning_rate": 9.675855021918869e-05, + "loss": 0.0591, + "step": 9570 + }, + { + "epoch": 0.2956972652632879, + "grad_norm": 0.38581374287605286, + "learning_rate": 9.674878222939561e-05, + "loss": 0.0588, + "step": 9580 + }, + { + "epoch": 0.2960059262917464, + "grad_norm": 0.4098140299320221, + "learning_rate": 9.673900003855681e-05, + "loss": 0.0574, + "step": 9590 + }, + { + "epoch": 0.29631458732020494, + "grad_norm": 0.28520286083221436, + "learning_rate": 9.672920364964389e-05, + "loss": 0.0574, + "step": 9600 + }, + { + "epoch": 0.2966232483486635, + "grad_norm": 0.39533358812332153, + "learning_rate": 9.671939306563269e-05, + "loss": 0.0554, + "step": 9610 + }, + { + "epoch": 0.29693190937712205, + "grad_norm": 0.9123843312263489, + "learning_rate": 9.670956828950345e-05, + "loss": 0.0595, + "step": 9620 + }, + { + "epoch": 0.2972405704055806, + "grad_norm": 0.24935217201709747, + "learning_rate": 9.669972932424065e-05, + "loss": 0.0614, + "step": 9630 + }, + { + "epoch": 0.29754923143403916, + "grad_norm": 0.6642357110977173, + "learning_rate": 9.668987617283312e-05, + "loss": 0.0596, + "step": 9640 + }, + { + "epoch": 0.29785789246249766, + "grad_norm": 0.42463093996047974, + "learning_rate": 9.668000883827397e-05, + "loss": 0.0609, + "step": 9650 + }, + { + "epoch": 0.2981665534909562, + "grad_norm": 0.3510684669017792, + "learning_rate": 9.667012732356067e-05, + "loss": 0.0539, + "step": 9660 + }, + { + "epoch": 0.2984752145194148, + "grad_norm": 0.35620301961898804, + "learning_rate": 9.666023163169493e-05, + "loss": 0.0574, + "step": 9670 + }, + { + "epoch": 0.29878387554787333, + "grad_norm": 0.3065052926540375, + "learning_rate": 9.665032176568281e-05, + "loss": 0.0567, + "step": 9680 + }, + { + "epoch": 0.2990925365763319, + "grad_norm": 0.3560742139816284, + "learning_rate": 9.664039772853469e-05, + "loss": 0.0562, + "step": 9690 + }, + { + "epoch": 0.2994011976047904, + "grad_norm": 0.33416616916656494, + "learning_rate": 9.663045952326518e-05, + "loss": 0.0549, + "step": 9700 + }, + { + "epoch": 0.29970985863324895, + "grad_norm": 0.2670486867427826, + "learning_rate": 9.662050715289328e-05, + "loss": 0.0586, + "step": 9710 + }, + { + "epoch": 0.3000185196617075, + "grad_norm": 0.4903070628643036, + "learning_rate": 9.661054062044226e-05, + "loss": 0.0607, + "step": 9720 + }, + { + "epoch": 0.30032718069016606, + "grad_norm": 0.30885180830955505, + "learning_rate": 9.660055992893968e-05, + "loss": 0.0562, + "step": 9730 + }, + { + "epoch": 0.3006358417186246, + "grad_norm": 0.2766093909740448, + "learning_rate": 9.659056508141739e-05, + "loss": 0.0654, + "step": 9740 + }, + { + "epoch": 0.3009445027470832, + "grad_norm": 0.2574992775917053, + "learning_rate": 9.658055608091161e-05, + "loss": 0.0613, + "step": 9750 + }, + { + "epoch": 0.3012531637755417, + "grad_norm": 0.5917277932167053, + "learning_rate": 9.657053293046276e-05, + "loss": 0.0556, + "step": 9760 + }, + { + "epoch": 0.30156182480400023, + "grad_norm": 0.33878570795059204, + "learning_rate": 9.656049563311564e-05, + "loss": 0.0577, + "step": 9770 + }, + { + "epoch": 0.3018704858324588, + "grad_norm": 0.31059691309928894, + "learning_rate": 9.655044419191929e-05, + "loss": 0.0602, + "step": 9780 + }, + { + "epoch": 0.30217914686091735, + "grad_norm": 0.2480134218931198, + "learning_rate": 9.654037860992711e-05, + "loss": 0.0579, + "step": 9790 + }, + { + "epoch": 0.3024878078893759, + "grad_norm": 0.3073379099369049, + "learning_rate": 9.653029889019672e-05, + "loss": 0.0529, + "step": 9800 + }, + { + "epoch": 0.30279646891783446, + "grad_norm": 0.3552672564983368, + "learning_rate": 9.65202050357901e-05, + "loss": 0.0528, + "step": 9810 + }, + { + "epoch": 0.30310512994629296, + "grad_norm": 0.27512964606285095, + "learning_rate": 9.651009704977347e-05, + "loss": 0.0573, + "step": 9820 + }, + { + "epoch": 0.3034137909747515, + "grad_norm": 0.6001698970794678, + "learning_rate": 9.649997493521738e-05, + "loss": 0.0559, + "step": 9830 + }, + { + "epoch": 0.3037224520032101, + "grad_norm": 0.3800532817840576, + "learning_rate": 9.64898386951967e-05, + "loss": 0.0587, + "step": 9840 + }, + { + "epoch": 0.30403111303166863, + "grad_norm": 0.2589397430419922, + "learning_rate": 9.647968833279049e-05, + "loss": 0.0555, + "step": 9850 + }, + { + "epoch": 0.3043397740601272, + "grad_norm": 0.3697051405906677, + "learning_rate": 9.646952385108218e-05, + "loss": 0.0553, + "step": 9860 + }, + { + "epoch": 0.3046484350885857, + "grad_norm": 0.2999069094657898, + "learning_rate": 9.645934525315951e-05, + "loss": 0.0586, + "step": 9870 + }, + { + "epoch": 0.30495709611704425, + "grad_norm": 0.33061277866363525, + "learning_rate": 9.644915254211442e-05, + "loss": 0.0594, + "step": 9880 + }, + { + "epoch": 0.3052657571455028, + "grad_norm": 0.337782621383667, + "learning_rate": 9.643894572104321e-05, + "loss": 0.0607, + "step": 9890 + }, + { + "epoch": 0.30557441817396136, + "grad_norm": 0.41777727007865906, + "learning_rate": 9.642872479304644e-05, + "loss": 0.0581, + "step": 9900 + }, + { + "epoch": 0.3058830792024199, + "grad_norm": 0.31907233595848083, + "learning_rate": 9.641848976122895e-05, + "loss": 0.0581, + "step": 9910 + }, + { + "epoch": 0.3061917402308785, + "grad_norm": 0.2468823790550232, + "learning_rate": 9.64082406286999e-05, + "loss": 0.0572, + "step": 9920 + }, + { + "epoch": 0.306500401259337, + "grad_norm": 0.3297012746334076, + "learning_rate": 9.639797739857269e-05, + "loss": 0.064, + "step": 9930 + }, + { + "epoch": 0.30680906228779553, + "grad_norm": 0.31968045234680176, + "learning_rate": 9.638770007396498e-05, + "loss": 0.058, + "step": 9940 + }, + { + "epoch": 0.3071177233162541, + "grad_norm": 0.28429409861564636, + "learning_rate": 9.63774086579988e-05, + "loss": 0.0605, + "step": 9950 + }, + { + "epoch": 0.30742638434471264, + "grad_norm": 0.29059791564941406, + "learning_rate": 9.63671031538004e-05, + "loss": 0.0571, + "step": 9960 + }, + { + "epoch": 0.3077350453731712, + "grad_norm": 1.0270791053771973, + "learning_rate": 9.635678356450031e-05, + "loss": 0.0576, + "step": 9970 + }, + { + "epoch": 0.30804370640162976, + "grad_norm": 0.7294179797172546, + "learning_rate": 9.634644989323336e-05, + "loss": 0.0617, + "step": 9980 + }, + { + "epoch": 0.30835236743008826, + "grad_norm": 0.3043498694896698, + "learning_rate": 9.633610214313861e-05, + "loss": 0.0577, + "step": 9990 + }, + { + "epoch": 0.3086610284585468, + "grad_norm": 0.3790475130081177, + "learning_rate": 9.632574031735951e-05, + "loss": 0.0569, + "step": 10000 + }, + { + "epoch": 0.3089696894870054, + "grad_norm": 0.2908840775489807, + "learning_rate": 9.631536441904364e-05, + "loss": 0.0526, + "step": 10010 + }, + { + "epoch": 0.30927835051546393, + "grad_norm": 0.47884243726730347, + "learning_rate": 9.630497445134293e-05, + "loss": 0.0588, + "step": 10020 + }, + { + "epoch": 0.3095870115439225, + "grad_norm": 0.38684043288230896, + "learning_rate": 9.62945704174136e-05, + "loss": 0.0555, + "step": 10030 + }, + { + "epoch": 0.309895672572381, + "grad_norm": 0.24158857762813568, + "learning_rate": 9.628415232041612e-05, + "loss": 0.0542, + "step": 10040 + }, + { + "epoch": 0.31020433360083954, + "grad_norm": 0.32554012537002563, + "learning_rate": 9.627372016351524e-05, + "loss": 0.0574, + "step": 10050 + }, + { + "epoch": 0.3105129946292981, + "grad_norm": 0.3084346055984497, + "learning_rate": 9.626327394987995e-05, + "loss": 0.0548, + "step": 10060 + }, + { + "epoch": 0.31082165565775666, + "grad_norm": 0.32617729902267456, + "learning_rate": 9.625281368268355e-05, + "loss": 0.0571, + "step": 10070 + }, + { + "epoch": 0.3111303166862152, + "grad_norm": 0.2933832108974457, + "learning_rate": 9.624233936510357e-05, + "loss": 0.0581, + "step": 10080 + }, + { + "epoch": 0.31143897771467377, + "grad_norm": 0.3968396484851837, + "learning_rate": 9.623185100032187e-05, + "loss": 0.056, + "step": 10090 + }, + { + "epoch": 0.31174763874313227, + "grad_norm": 0.3545081615447998, + "learning_rate": 9.62213485915245e-05, + "loss": 0.0586, + "step": 10100 + }, + { + "epoch": 0.31205629977159083, + "grad_norm": 0.37151849269866943, + "learning_rate": 9.621083214190186e-05, + "loss": 0.0574, + "step": 10110 + }, + { + "epoch": 0.3123649608000494, + "grad_norm": 0.3900144398212433, + "learning_rate": 9.62003016546485e-05, + "loss": 0.0569, + "step": 10120 + }, + { + "epoch": 0.31267362182850794, + "grad_norm": 0.47107046842575073, + "learning_rate": 9.618975713296339e-05, + "loss": 0.0549, + "step": 10130 + }, + { + "epoch": 0.3129822828569665, + "grad_norm": 0.35863035917282104, + "learning_rate": 9.61791985800496e-05, + "loss": 0.0573, + "step": 10140 + }, + { + "epoch": 0.313290943885425, + "grad_norm": 0.4012429118156433, + "learning_rate": 9.616862599911458e-05, + "loss": 0.0567, + "step": 10150 + }, + { + "epoch": 0.31359960491388356, + "grad_norm": 0.3796987533569336, + "learning_rate": 9.615803939337e-05, + "loss": 0.0586, + "step": 10160 + }, + { + "epoch": 0.3139082659423421, + "grad_norm": 0.4267863631248474, + "learning_rate": 9.614743876603178e-05, + "loss": 0.0628, + "step": 10170 + }, + { + "epoch": 0.31421692697080067, + "grad_norm": 0.38627779483795166, + "learning_rate": 9.613682412032013e-05, + "loss": 0.0603, + "step": 10180 + }, + { + "epoch": 0.3145255879992592, + "grad_norm": 0.3259195387363434, + "learning_rate": 9.612619545945947e-05, + "loss": 0.0548, + "step": 10190 + }, + { + "epoch": 0.3148342490277178, + "grad_norm": 0.29857635498046875, + "learning_rate": 9.611555278667852e-05, + "loss": 0.0544, + "step": 10200 + }, + { + "epoch": 0.3151429100561763, + "grad_norm": 0.33251410722732544, + "learning_rate": 9.610489610521024e-05, + "loss": 0.0578, + "step": 10210 + }, + { + "epoch": 0.31545157108463484, + "grad_norm": 0.3442172408103943, + "learning_rate": 9.609422541829187e-05, + "loss": 0.0604, + "step": 10220 + }, + { + "epoch": 0.3157602321130934, + "grad_norm": 0.5024712085723877, + "learning_rate": 9.608354072916486e-05, + "loss": 0.0521, + "step": 10230 + }, + { + "epoch": 0.31606889314155195, + "grad_norm": 0.40970703959465027, + "learning_rate": 9.607284204107493e-05, + "loss": 0.0509, + "step": 10240 + }, + { + "epoch": 0.3163775541700105, + "grad_norm": 0.38036295771598816, + "learning_rate": 9.606212935727208e-05, + "loss": 0.0582, + "step": 10250 + }, + { + "epoch": 0.31668621519846907, + "grad_norm": 0.3950102925300598, + "learning_rate": 9.605140268101052e-05, + "loss": 0.0614, + "step": 10260 + }, + { + "epoch": 0.31699487622692757, + "grad_norm": 0.3359011709690094, + "learning_rate": 9.604066201554875e-05, + "loss": 0.0572, + "step": 10270 + }, + { + "epoch": 0.3173035372553861, + "grad_norm": 0.35198912024497986, + "learning_rate": 9.60299073641495e-05, + "loss": 0.0605, + "step": 10280 + }, + { + "epoch": 0.3176121982838447, + "grad_norm": 0.33616045117378235, + "learning_rate": 9.601913873007974e-05, + "loss": 0.0588, + "step": 10290 + }, + { + "epoch": 0.31792085931230324, + "grad_norm": 0.40930813550949097, + "learning_rate": 9.60083561166107e-05, + "loss": 0.059, + "step": 10300 + }, + { + "epoch": 0.3182295203407618, + "grad_norm": 0.3550591468811035, + "learning_rate": 9.599755952701783e-05, + "loss": 0.0558, + "step": 10310 + }, + { + "epoch": 0.3185381813692203, + "grad_norm": 0.299388587474823, + "learning_rate": 9.598674896458089e-05, + "loss": 0.0554, + "step": 10320 + }, + { + "epoch": 0.31884684239767885, + "grad_norm": 0.3866835832595825, + "learning_rate": 9.597592443258383e-05, + "loss": 0.0588, + "step": 10330 + }, + { + "epoch": 0.3191555034261374, + "grad_norm": 0.29153972864151, + "learning_rate": 9.596508593431483e-05, + "loss": 0.0586, + "step": 10340 + }, + { + "epoch": 0.31946416445459597, + "grad_norm": 0.34436100721359253, + "learning_rate": 9.59542334730664e-05, + "loss": 0.0602, + "step": 10350 + }, + { + "epoch": 0.3197728254830545, + "grad_norm": 0.4862590432167053, + "learning_rate": 9.594336705213516e-05, + "loss": 0.0558, + "step": 10360 + }, + { + "epoch": 0.3200814865115131, + "grad_norm": 0.38206717371940613, + "learning_rate": 9.593248667482208e-05, + "loss": 0.0562, + "step": 10370 + }, + { + "epoch": 0.3203901475399716, + "grad_norm": 0.3123026490211487, + "learning_rate": 9.592159234443233e-05, + "loss": 0.0543, + "step": 10380 + }, + { + "epoch": 0.32069880856843014, + "grad_norm": 0.29053932428359985, + "learning_rate": 9.59106840642753e-05, + "loss": 0.0614, + "step": 10390 + }, + { + "epoch": 0.3210074695968887, + "grad_norm": 0.336548388004303, + "learning_rate": 9.589976183766467e-05, + "loss": 0.0603, + "step": 10400 + }, + { + "epoch": 0.32131613062534725, + "grad_norm": 0.4339388608932495, + "learning_rate": 9.58888256679183e-05, + "loss": 0.0582, + "step": 10410 + }, + { + "epoch": 0.3216247916538058, + "grad_norm": 0.3406897485256195, + "learning_rate": 9.587787555835832e-05, + "loss": 0.0542, + "step": 10420 + }, + { + "epoch": 0.32193345268226436, + "grad_norm": 0.35645410418510437, + "learning_rate": 9.586691151231107e-05, + "loss": 0.053, + "step": 10430 + }, + { + "epoch": 0.32224211371072287, + "grad_norm": 0.3719894289970398, + "learning_rate": 9.585593353310715e-05, + "loss": 0.0562, + "step": 10440 + }, + { + "epoch": 0.3225507747391814, + "grad_norm": 0.3469242453575134, + "learning_rate": 9.58449416240814e-05, + "loss": 0.0546, + "step": 10450 + }, + { + "epoch": 0.32285943576764, + "grad_norm": 0.282184362411499, + "learning_rate": 9.583393578857283e-05, + "loss": 0.0563, + "step": 10460 + }, + { + "epoch": 0.32316809679609854, + "grad_norm": 0.2938305139541626, + "learning_rate": 9.582291602992474e-05, + "loss": 0.0585, + "step": 10470 + }, + { + "epoch": 0.3234767578245571, + "grad_norm": 0.3547019362449646, + "learning_rate": 9.581188235148466e-05, + "loss": 0.0573, + "step": 10480 + }, + { + "epoch": 0.3237854188530156, + "grad_norm": 0.5002971887588501, + "learning_rate": 9.58008347566043e-05, + "loss": 0.0589, + "step": 10490 + }, + { + "epoch": 0.32409407988147415, + "grad_norm": 0.3778490424156189, + "learning_rate": 9.578977324863965e-05, + "loss": 0.062, + "step": 10500 + }, + { + "epoch": 0.3244027409099327, + "grad_norm": 0.5657587647438049, + "learning_rate": 9.577869783095089e-05, + "loss": 0.0548, + "step": 10510 + }, + { + "epoch": 0.32471140193839126, + "grad_norm": 0.2923542261123657, + "learning_rate": 9.576760850690245e-05, + "loss": 0.0584, + "step": 10520 + }, + { + "epoch": 0.3250200629668498, + "grad_norm": 0.3465745151042938, + "learning_rate": 9.575650527986298e-05, + "loss": 0.059, + "step": 10530 + }, + { + "epoch": 0.3253287239953084, + "grad_norm": 0.29953595995903015, + "learning_rate": 9.574538815320531e-05, + "loss": 0.0596, + "step": 10540 + }, + { + "epoch": 0.3256373850237669, + "grad_norm": 0.4019628167152405, + "learning_rate": 9.573425713030656e-05, + "loss": 0.0576, + "step": 10550 + }, + { + "epoch": 0.32594604605222544, + "grad_norm": 0.4151754379272461, + "learning_rate": 9.572311221454806e-05, + "loss": 0.0572, + "step": 10560 + }, + { + "epoch": 0.326254707080684, + "grad_norm": 0.3527437150478363, + "learning_rate": 9.57119534093153e-05, + "loss": 0.0589, + "step": 10570 + }, + { + "epoch": 0.32656336810914255, + "grad_norm": 0.3897445797920227, + "learning_rate": 9.570078071799806e-05, + "loss": 0.0592, + "step": 10580 + }, + { + "epoch": 0.3268720291376011, + "grad_norm": 0.328402578830719, + "learning_rate": 9.568959414399028e-05, + "loss": 0.0578, + "step": 10590 + }, + { + "epoch": 0.3271806901660596, + "grad_norm": 0.4121093153953552, + "learning_rate": 9.567839369069018e-05, + "loss": 0.0554, + "step": 10600 + }, + { + "epoch": 0.32748935119451816, + "grad_norm": 0.3247385323047638, + "learning_rate": 9.566717936150013e-05, + "loss": 0.0549, + "step": 10610 + }, + { + "epoch": 0.3277980122229767, + "grad_norm": 0.26147130131721497, + "learning_rate": 9.565595115982678e-05, + "loss": 0.061, + "step": 10620 + }, + { + "epoch": 0.3281066732514353, + "grad_norm": 0.4329439103603363, + "learning_rate": 9.564470908908094e-05, + "loss": 0.0568, + "step": 10630 + }, + { + "epoch": 0.32841533427989383, + "grad_norm": 0.33028659224510193, + "learning_rate": 9.563345315267764e-05, + "loss": 0.0623, + "step": 10640 + }, + { + "epoch": 0.3287239953083524, + "grad_norm": 0.377162367105484, + "learning_rate": 9.562218335403616e-05, + "loss": 0.0637, + "step": 10650 + }, + { + "epoch": 0.3290326563368109, + "grad_norm": 0.33241701126098633, + "learning_rate": 9.561089969657999e-05, + "loss": 0.0515, + "step": 10660 + }, + { + "epoch": 0.32934131736526945, + "grad_norm": 0.24740980565547943, + "learning_rate": 9.559960218373673e-05, + "loss": 0.0571, + "step": 10670 + }, + { + "epoch": 0.329649978393728, + "grad_norm": 0.30073872208595276, + "learning_rate": 9.558829081893836e-05, + "loss": 0.0591, + "step": 10680 + }, + { + "epoch": 0.32995863942218656, + "grad_norm": 0.4204317033290863, + "learning_rate": 9.55769656056209e-05, + "loss": 0.0599, + "step": 10690 + }, + { + "epoch": 0.3302673004506451, + "grad_norm": 0.3687698543071747, + "learning_rate": 9.556562654722469e-05, + "loss": 0.0627, + "step": 10700 + }, + { + "epoch": 0.3305759614791037, + "grad_norm": 0.3176354467868805, + "learning_rate": 9.555427364719422e-05, + "loss": 0.0558, + "step": 10710 + }, + { + "epoch": 0.3308846225075622, + "grad_norm": 0.3752564787864685, + "learning_rate": 9.55429069089782e-05, + "loss": 0.0597, + "step": 10720 + }, + { + "epoch": 0.33119328353602073, + "grad_norm": 0.466239333152771, + "learning_rate": 9.553152633602956e-05, + "loss": 0.0593, + "step": 10730 + }, + { + "epoch": 0.3315019445644793, + "grad_norm": 0.36478370428085327, + "learning_rate": 9.552013193180543e-05, + "loss": 0.0598, + "step": 10740 + }, + { + "epoch": 0.33181060559293785, + "grad_norm": 0.3458446264266968, + "learning_rate": 9.550872369976707e-05, + "loss": 0.058, + "step": 10750 + }, + { + "epoch": 0.3321192666213964, + "grad_norm": 0.37919896841049194, + "learning_rate": 9.549730164338007e-05, + "loss": 0.0556, + "step": 10760 + }, + { + "epoch": 0.3324279276498549, + "grad_norm": 0.31846168637275696, + "learning_rate": 9.548586576611408e-05, + "loss": 0.0536, + "step": 10770 + }, + { + "epoch": 0.33273658867831346, + "grad_norm": 0.4319573938846588, + "learning_rate": 9.54744160714431e-05, + "loss": 0.0545, + "step": 10780 + }, + { + "epoch": 0.333045249706772, + "grad_norm": 0.2695809602737427, + "learning_rate": 9.546295256284516e-05, + "loss": 0.0581, + "step": 10790 + }, + { + "epoch": 0.3333539107352306, + "grad_norm": 0.35071423649787903, + "learning_rate": 9.545147524380265e-05, + "loss": 0.06, + "step": 10800 + }, + { + "epoch": 0.33366257176368913, + "grad_norm": 0.4055568277835846, + "learning_rate": 9.543998411780201e-05, + "loss": 0.0588, + "step": 10810 + }, + { + "epoch": 0.3339712327921477, + "grad_norm": 0.3292250335216522, + "learning_rate": 9.542847918833397e-05, + "loss": 0.0571, + "step": 10820 + }, + { + "epoch": 0.3342798938206062, + "grad_norm": 0.3105493187904358, + "learning_rate": 9.541696045889343e-05, + "loss": 0.0546, + "step": 10830 + }, + { + "epoch": 0.33458855484906475, + "grad_norm": 0.40082359313964844, + "learning_rate": 9.540542793297947e-05, + "loss": 0.055, + "step": 10840 + }, + { + "epoch": 0.3348972158775233, + "grad_norm": 0.436132550239563, + "learning_rate": 9.539388161409537e-05, + "loss": 0.0606, + "step": 10850 + }, + { + "epoch": 0.33520587690598186, + "grad_norm": 0.3884340822696686, + "learning_rate": 9.538232150574857e-05, + "loss": 0.055, + "step": 10860 + }, + { + "epoch": 0.3355145379344404, + "grad_norm": 0.2412654310464859, + "learning_rate": 9.537074761145076e-05, + "loss": 0.0566, + "step": 10870 + }, + { + "epoch": 0.3358231989628989, + "grad_norm": 0.3310239911079407, + "learning_rate": 9.535915993471778e-05, + "loss": 0.0541, + "step": 10880 + }, + { + "epoch": 0.3361318599913575, + "grad_norm": 0.3592749238014221, + "learning_rate": 9.534755847906964e-05, + "loss": 0.0533, + "step": 10890 + }, + { + "epoch": 0.33644052101981603, + "grad_norm": 0.2833961546421051, + "learning_rate": 9.533594324803057e-05, + "loss": 0.0562, + "step": 10900 + }, + { + "epoch": 0.3367491820482746, + "grad_norm": 0.3398839235305786, + "learning_rate": 9.532431424512895e-05, + "loss": 0.0573, + "step": 10910 + }, + { + "epoch": 0.33705784307673314, + "grad_norm": 0.2873094975948334, + "learning_rate": 9.531267147389741e-05, + "loss": 0.058, + "step": 10920 + }, + { + "epoch": 0.3373665041051917, + "grad_norm": 0.47003424167633057, + "learning_rate": 9.530101493787266e-05, + "loss": 0.0606, + "step": 10930 + }, + { + "epoch": 0.3376751651336502, + "grad_norm": 0.33040034770965576, + "learning_rate": 9.528934464059571e-05, + "loss": 0.0571, + "step": 10940 + }, + { + "epoch": 0.33798382616210876, + "grad_norm": 0.3516250550746918, + "learning_rate": 9.527766058561163e-05, + "loss": 0.0547, + "step": 10950 + }, + { + "epoch": 0.3382924871905673, + "grad_norm": 0.3590119481086731, + "learning_rate": 9.526596277646976e-05, + "loss": 0.0548, + "step": 10960 + }, + { + "epoch": 0.33860114821902587, + "grad_norm": 0.25238916277885437, + "learning_rate": 9.525425121672358e-05, + "loss": 0.0631, + "step": 10970 + }, + { + "epoch": 0.33890980924748443, + "grad_norm": 0.41489744186401367, + "learning_rate": 9.524252590993074e-05, + "loss": 0.0547, + "step": 10980 + }, + { + "epoch": 0.339218470275943, + "grad_norm": 0.3024161159992218, + "learning_rate": 9.523078685965309e-05, + "loss": 0.0601, + "step": 10990 + }, + { + "epoch": 0.3395271313044015, + "grad_norm": 0.4563504159450531, + "learning_rate": 9.521903406945664e-05, + "loss": 0.0543, + "step": 11000 + }, + { + "epoch": 0.33983579233286004, + "grad_norm": 0.39671921730041504, + "learning_rate": 9.520726754291158e-05, + "loss": 0.0562, + "step": 11010 + }, + { + "epoch": 0.3401444533613186, + "grad_norm": 0.4110635221004486, + "learning_rate": 9.519548728359227e-05, + "loss": 0.0589, + "step": 11020 + }, + { + "epoch": 0.34045311438977716, + "grad_norm": 0.32278770208358765, + "learning_rate": 9.518369329507726e-05, + "loss": 0.051, + "step": 11030 + }, + { + "epoch": 0.3407617754182357, + "grad_norm": 0.37574127316474915, + "learning_rate": 9.51718855809492e-05, + "loss": 0.0537, + "step": 11040 + }, + { + "epoch": 0.3410704364466942, + "grad_norm": 0.3045138120651245, + "learning_rate": 9.516006414479502e-05, + "loss": 0.0569, + "step": 11050 + }, + { + "epoch": 0.34137909747515277, + "grad_norm": 0.25207820534706116, + "learning_rate": 9.514822899020572e-05, + "loss": 0.0555, + "step": 11060 + }, + { + "epoch": 0.3416877585036113, + "grad_norm": 0.29355520009994507, + "learning_rate": 9.513638012077654e-05, + "loss": 0.0513, + "step": 11070 + }, + { + "epoch": 0.3419964195320699, + "grad_norm": 0.40934666991233826, + "learning_rate": 9.512451754010683e-05, + "loss": 0.0589, + "step": 11080 + }, + { + "epoch": 0.34230508056052844, + "grad_norm": 0.34146344661712646, + "learning_rate": 9.511264125180013e-05, + "loss": 0.054, + "step": 11090 + }, + { + "epoch": 0.342613741588987, + "grad_norm": 0.29882925748825073, + "learning_rate": 9.510075125946414e-05, + "loss": 0.0553, + "step": 11100 + }, + { + "epoch": 0.3429224026174455, + "grad_norm": 0.41532599925994873, + "learning_rate": 9.508884756671075e-05, + "loss": 0.0595, + "step": 11110 + }, + { + "epoch": 0.34323106364590406, + "grad_norm": 0.31476572155952454, + "learning_rate": 9.507693017715596e-05, + "loss": 0.0558, + "step": 11120 + }, + { + "epoch": 0.3435397246743626, + "grad_norm": 0.3481561839580536, + "learning_rate": 9.506499909441997e-05, + "loss": 0.0589, + "step": 11130 + }, + { + "epoch": 0.34384838570282117, + "grad_norm": 0.37436285614967346, + "learning_rate": 9.505305432212713e-05, + "loss": 0.0549, + "step": 11140 + }, + { + "epoch": 0.3441570467312797, + "grad_norm": 0.27179038524627686, + "learning_rate": 9.504109586390595e-05, + "loss": 0.0509, + "step": 11150 + }, + { + "epoch": 0.3444657077597383, + "grad_norm": 0.3654322624206543, + "learning_rate": 9.502912372338908e-05, + "loss": 0.0588, + "step": 11160 + }, + { + "epoch": 0.3447743687881968, + "grad_norm": 0.3180154860019684, + "learning_rate": 9.501713790421335e-05, + "loss": 0.0553, + "step": 11170 + }, + { + "epoch": 0.34508302981665534, + "grad_norm": 0.40438205003738403, + "learning_rate": 9.500513841001974e-05, + "loss": 0.0512, + "step": 11180 + }, + { + "epoch": 0.3453916908451139, + "grad_norm": 0.4950290322303772, + "learning_rate": 9.499312524445336e-05, + "loss": 0.0628, + "step": 11190 + }, + { + "epoch": 0.34570035187357245, + "grad_norm": 0.3928847312927246, + "learning_rate": 9.498109841116351e-05, + "loss": 0.057, + "step": 11200 + }, + { + "epoch": 0.346009012902031, + "grad_norm": 0.4741062521934509, + "learning_rate": 9.496905791380363e-05, + "loss": 0.0595, + "step": 11210 + }, + { + "epoch": 0.3463176739304895, + "grad_norm": 0.28509292006492615, + "learning_rate": 9.495700375603129e-05, + "loss": 0.0528, + "step": 11220 + }, + { + "epoch": 0.34662633495894807, + "grad_norm": 0.22448532283306122, + "learning_rate": 9.494493594150822e-05, + "loss": 0.0609, + "step": 11230 + }, + { + "epoch": 0.3469349959874066, + "grad_norm": 0.3372233808040619, + "learning_rate": 9.493285447390032e-05, + "loss": 0.0572, + "step": 11240 + }, + { + "epoch": 0.3472436570158652, + "grad_norm": 0.27831345796585083, + "learning_rate": 9.492075935687761e-05, + "loss": 0.0552, + "step": 11250 + }, + { + "epoch": 0.34755231804432374, + "grad_norm": 0.24344338476657867, + "learning_rate": 9.490865059411427e-05, + "loss": 0.0546, + "step": 11260 + }, + { + "epoch": 0.3478609790727823, + "grad_norm": 0.37084490060806274, + "learning_rate": 9.489652818928863e-05, + "loss": 0.0552, + "step": 11270 + }, + { + "epoch": 0.3481696401012408, + "grad_norm": 0.26783257722854614, + "learning_rate": 9.488439214608315e-05, + "loss": 0.0506, + "step": 11280 + }, + { + "epoch": 0.34847830112969935, + "grad_norm": 0.290149450302124, + "learning_rate": 9.487224246818444e-05, + "loss": 0.0549, + "step": 11290 + }, + { + "epoch": 0.3487869621581579, + "grad_norm": 0.31643521785736084, + "learning_rate": 9.486007915928325e-05, + "loss": 0.054, + "step": 11300 + }, + { + "epoch": 0.34909562318661647, + "grad_norm": 0.3481188118457794, + "learning_rate": 9.484790222307448e-05, + "loss": 0.0523, + "step": 11310 + }, + { + "epoch": 0.349404284215075, + "grad_norm": 0.3637100160121918, + "learning_rate": 9.483571166325716e-05, + "loss": 0.0567, + "step": 11320 + }, + { + "epoch": 0.3497129452435335, + "grad_norm": 0.41394728422164917, + "learning_rate": 9.482350748353444e-05, + "loss": 0.0581, + "step": 11330 + }, + { + "epoch": 0.3500216062719921, + "grad_norm": 0.2815488576889038, + "learning_rate": 9.481128968761363e-05, + "loss": 0.0521, + "step": 11340 + }, + { + "epoch": 0.35033026730045064, + "grad_norm": 0.31432586908340454, + "learning_rate": 9.479905827920621e-05, + "loss": 0.0597, + "step": 11350 + }, + { + "epoch": 0.3506389283289092, + "grad_norm": 0.3692314922809601, + "learning_rate": 9.478681326202773e-05, + "loss": 0.0561, + "step": 11360 + }, + { + "epoch": 0.35094758935736775, + "grad_norm": 0.3536187708377838, + "learning_rate": 9.477455463979791e-05, + "loss": 0.0574, + "step": 11370 + }, + { + "epoch": 0.3512562503858263, + "grad_norm": 0.39141160249710083, + "learning_rate": 9.476228241624059e-05, + "loss": 0.0565, + "step": 11380 + }, + { + "epoch": 0.3515649114142848, + "grad_norm": 0.3613574504852295, + "learning_rate": 9.474999659508374e-05, + "loss": 0.0592, + "step": 11390 + }, + { + "epoch": 0.35187357244274337, + "grad_norm": 0.41762179136276245, + "learning_rate": 9.47376971800595e-05, + "loss": 0.0606, + "step": 11400 + }, + { + "epoch": 0.3521822334712019, + "grad_norm": 0.3608429431915283, + "learning_rate": 9.472538417490409e-05, + "loss": 0.0564, + "step": 11410 + }, + { + "epoch": 0.3524908944996605, + "grad_norm": 0.2737177312374115, + "learning_rate": 9.471305758335784e-05, + "loss": 0.0546, + "step": 11420 + }, + { + "epoch": 0.35279955552811904, + "grad_norm": 0.4083748757839203, + "learning_rate": 9.47007174091653e-05, + "loss": 0.057, + "step": 11430 + }, + { + "epoch": 0.3531082165565776, + "grad_norm": 0.3651997745037079, + "learning_rate": 9.468836365607507e-05, + "loss": 0.0532, + "step": 11440 + }, + { + "epoch": 0.3534168775850361, + "grad_norm": 0.5031157732009888, + "learning_rate": 9.467599632783988e-05, + "loss": 0.0567, + "step": 11450 + }, + { + "epoch": 0.35372553861349465, + "grad_norm": 0.2984493374824524, + "learning_rate": 9.466361542821662e-05, + "loss": 0.0589, + "step": 11460 + }, + { + "epoch": 0.3540341996419532, + "grad_norm": 0.32906684279441833, + "learning_rate": 9.465122096096625e-05, + "loss": 0.0528, + "step": 11470 + }, + { + "epoch": 0.35434286067041176, + "grad_norm": 0.30348944664001465, + "learning_rate": 9.463881292985391e-05, + "loss": 0.0551, + "step": 11480 + }, + { + "epoch": 0.3546515216988703, + "grad_norm": 0.37093526124954224, + "learning_rate": 9.462639133864881e-05, + "loss": 0.0567, + "step": 11490 + }, + { + "epoch": 0.3549601827273288, + "grad_norm": 0.3958415985107422, + "learning_rate": 9.461395619112432e-05, + "loss": 0.0549, + "step": 11500 + }, + { + "epoch": 0.3552688437557874, + "grad_norm": 0.3348768353462219, + "learning_rate": 9.460150749105791e-05, + "loss": 0.0603, + "step": 11510 + }, + { + "epoch": 0.35557750478424593, + "grad_norm": 0.4638233482837677, + "learning_rate": 9.458904524223116e-05, + "loss": 0.061, + "step": 11520 + }, + { + "epoch": 0.3558861658127045, + "grad_norm": 0.298147976398468, + "learning_rate": 9.457656944842976e-05, + "loss": 0.0631, + "step": 11530 + }, + { + "epoch": 0.35619482684116305, + "grad_norm": 0.3168477416038513, + "learning_rate": 9.456408011344353e-05, + "loss": 0.06, + "step": 11540 + }, + { + "epoch": 0.3565034878696216, + "grad_norm": 0.3983292281627655, + "learning_rate": 9.455157724106643e-05, + "loss": 0.0623, + "step": 11550 + }, + { + "epoch": 0.3568121488980801, + "grad_norm": 0.3427877426147461, + "learning_rate": 9.453906083509647e-05, + "loss": 0.0591, + "step": 11560 + }, + { + "epoch": 0.35712080992653866, + "grad_norm": 0.37151047587394714, + "learning_rate": 9.45265308993358e-05, + "loss": 0.0537, + "step": 11570 + }, + { + "epoch": 0.3574294709549972, + "grad_norm": 0.2852877974510193, + "learning_rate": 9.451398743759071e-05, + "loss": 0.0575, + "step": 11580 + }, + { + "epoch": 0.3577381319834558, + "grad_norm": 0.42948299646377563, + "learning_rate": 9.450143045367156e-05, + "loss": 0.0579, + "step": 11590 + }, + { + "epoch": 0.35804679301191433, + "grad_norm": 0.4042137861251831, + "learning_rate": 9.448885995139283e-05, + "loss": 0.0578, + "step": 11600 + }, + { + "epoch": 0.3583554540403729, + "grad_norm": 0.5550439953804016, + "learning_rate": 9.44762759345731e-05, + "loss": 0.0613, + "step": 11610 + }, + { + "epoch": 0.3586641150688314, + "grad_norm": 0.3205573558807373, + "learning_rate": 9.446367840703509e-05, + "loss": 0.0624, + "step": 11620 + }, + { + "epoch": 0.35897277609728995, + "grad_norm": 0.3308817446231842, + "learning_rate": 9.445106737260556e-05, + "loss": 0.0658, + "step": 11630 + }, + { + "epoch": 0.3592814371257485, + "grad_norm": 0.32160484790802, + "learning_rate": 9.443844283511543e-05, + "loss": 0.0567, + "step": 11640 + }, + { + "epoch": 0.35959009815420706, + "grad_norm": 0.2794205844402313, + "learning_rate": 9.442580479839968e-05, + "loss": 0.0564, + "step": 11650 + }, + { + "epoch": 0.3598987591826656, + "grad_norm": 0.32878729701042175, + "learning_rate": 9.441315326629745e-05, + "loss": 0.0571, + "step": 11660 + }, + { + "epoch": 0.3602074202111241, + "grad_norm": 0.34618476033210754, + "learning_rate": 9.44004882426519e-05, + "loss": 0.0587, + "step": 11670 + }, + { + "epoch": 0.3605160812395827, + "grad_norm": 0.288684606552124, + "learning_rate": 9.438780973131037e-05, + "loss": 0.0522, + "step": 11680 + }, + { + "epoch": 0.36082474226804123, + "grad_norm": 0.30736732482910156, + "learning_rate": 9.437511773612423e-05, + "loss": 0.0539, + "step": 11690 + }, + { + "epoch": 0.3611334032964998, + "grad_norm": 0.30629685521125793, + "learning_rate": 9.436241226094896e-05, + "loss": 0.0566, + "step": 11700 + }, + { + "epoch": 0.36144206432495835, + "grad_norm": 0.2939259707927704, + "learning_rate": 9.434969330964418e-05, + "loss": 0.0542, + "step": 11710 + }, + { + "epoch": 0.3617507253534169, + "grad_norm": 0.533110499382019, + "learning_rate": 9.433696088607356e-05, + "loss": 0.057, + "step": 11720 + }, + { + "epoch": 0.3620593863818754, + "grad_norm": 0.2762441337108612, + "learning_rate": 9.432421499410486e-05, + "loss": 0.0563, + "step": 11730 + }, + { + "epoch": 0.36236804741033396, + "grad_norm": 0.4682613015174866, + "learning_rate": 9.431145563760998e-05, + "loss": 0.0614, + "step": 11740 + }, + { + "epoch": 0.3626767084387925, + "grad_norm": 0.3430090844631195, + "learning_rate": 9.429868282046484e-05, + "loss": 0.0573, + "step": 11750 + }, + { + "epoch": 0.3629853694672511, + "grad_norm": 0.32459309697151184, + "learning_rate": 9.428589654654951e-05, + "loss": 0.056, + "step": 11760 + }, + { + "epoch": 0.36329403049570963, + "grad_norm": 0.30021053552627563, + "learning_rate": 9.42730968197481e-05, + "loss": 0.0568, + "step": 11770 + }, + { + "epoch": 0.36360269152416813, + "grad_norm": 0.368662565946579, + "learning_rate": 9.426028364394883e-05, + "loss": 0.0517, + "step": 11780 + }, + { + "epoch": 0.3639113525526267, + "grad_norm": 0.3391352891921997, + "learning_rate": 9.424745702304402e-05, + "loss": 0.0556, + "step": 11790 + }, + { + "epoch": 0.36422001358108524, + "grad_norm": 0.4548531770706177, + "learning_rate": 9.423461696093006e-05, + "loss": 0.0623, + "step": 11800 + }, + { + "epoch": 0.3645286746095438, + "grad_norm": 0.284584641456604, + "learning_rate": 9.422176346150741e-05, + "loss": 0.0563, + "step": 11810 + }, + { + "epoch": 0.36483733563800236, + "grad_norm": 0.3187182545661926, + "learning_rate": 9.420889652868063e-05, + "loss": 0.0638, + "step": 11820 + }, + { + "epoch": 0.3651459966664609, + "grad_norm": 0.2912245988845825, + "learning_rate": 9.419601616635836e-05, + "loss": 0.0534, + "step": 11830 + }, + { + "epoch": 0.3654546576949194, + "grad_norm": 0.6853069067001343, + "learning_rate": 9.418312237845331e-05, + "loss": 0.0646, + "step": 11840 + }, + { + "epoch": 0.365763318723378, + "grad_norm": 0.3128475546836853, + "learning_rate": 9.417021516888225e-05, + "loss": 0.059, + "step": 11850 + }, + { + "epoch": 0.36607197975183653, + "grad_norm": 0.34369322657585144, + "learning_rate": 9.415729454156608e-05, + "loss": 0.055, + "step": 11860 + }, + { + "epoch": 0.3663806407802951, + "grad_norm": 0.28095903992652893, + "learning_rate": 9.414436050042973e-05, + "loss": 0.0549, + "step": 11870 + }, + { + "epoch": 0.36668930180875364, + "grad_norm": 0.2991349697113037, + "learning_rate": 9.413141304940223e-05, + "loss": 0.0609, + "step": 11880 + }, + { + "epoch": 0.3669979628372122, + "grad_norm": 0.3703649938106537, + "learning_rate": 9.411845219241666e-05, + "loss": 0.0573, + "step": 11890 + }, + { + "epoch": 0.3673066238656707, + "grad_norm": 0.5316898822784424, + "learning_rate": 9.410547793341021e-05, + "loss": 0.057, + "step": 11900 + }, + { + "epoch": 0.36761528489412926, + "grad_norm": 0.7578430771827698, + "learning_rate": 9.409249027632408e-05, + "loss": 0.0568, + "step": 11910 + }, + { + "epoch": 0.3679239459225878, + "grad_norm": 0.34560859203338623, + "learning_rate": 9.407948922510362e-05, + "loss": 0.054, + "step": 11920 + }, + { + "epoch": 0.36823260695104637, + "grad_norm": 0.45018765330314636, + "learning_rate": 9.406647478369817e-05, + "loss": 0.0656, + "step": 11930 + }, + { + "epoch": 0.3685412679795049, + "grad_norm": 0.39929404854774475, + "learning_rate": 9.405344695606118e-05, + "loss": 0.0594, + "step": 11940 + }, + { + "epoch": 0.36884992900796343, + "grad_norm": 0.3620223104953766, + "learning_rate": 9.404040574615018e-05, + "loss": 0.0523, + "step": 11950 + }, + { + "epoch": 0.369158590036422, + "grad_norm": 0.3766704797744751, + "learning_rate": 9.402735115792674e-05, + "loss": 0.0602, + "step": 11960 + }, + { + "epoch": 0.36946725106488054, + "grad_norm": 0.29182612895965576, + "learning_rate": 9.401428319535649e-05, + "loss": 0.0536, + "step": 11970 + }, + { + "epoch": 0.3697759120933391, + "grad_norm": 0.29785069823265076, + "learning_rate": 9.400120186240912e-05, + "loss": 0.058, + "step": 11980 + }, + { + "epoch": 0.37008457312179766, + "grad_norm": 0.7326500415802002, + "learning_rate": 9.398810716305844e-05, + "loss": 0.0564, + "step": 11990 + }, + { + "epoch": 0.3703932341502562, + "grad_norm": 0.2626456022262573, + "learning_rate": 9.397499910128222e-05, + "loss": 0.0593, + "step": 12000 + }, + { + "epoch": 0.3707018951787147, + "grad_norm": 0.35724231600761414, + "learning_rate": 9.396187768106237e-05, + "loss": 0.059, + "step": 12010 + }, + { + "epoch": 0.37101055620717327, + "grad_norm": 0.348435640335083, + "learning_rate": 9.394874290638482e-05, + "loss": 0.0527, + "step": 12020 + }, + { + "epoch": 0.3713192172356318, + "grad_norm": 0.31062057614326477, + "learning_rate": 9.393559478123959e-05, + "loss": 0.0573, + "step": 12030 + }, + { + "epoch": 0.3716278782640904, + "grad_norm": 0.4150077700614929, + "learning_rate": 9.39224333096207e-05, + "loss": 0.057, + "step": 12040 + }, + { + "epoch": 0.37193653929254894, + "grad_norm": 0.5122307538986206, + "learning_rate": 9.390925849552629e-05, + "loss": 0.0577, + "step": 12050 + }, + { + "epoch": 0.3722452003210075, + "grad_norm": 0.35102275013923645, + "learning_rate": 9.389607034295849e-05, + "loss": 0.0562, + "step": 12060 + }, + { + "epoch": 0.372553861349466, + "grad_norm": 0.28012609481811523, + "learning_rate": 9.388286885592355e-05, + "loss": 0.0542, + "step": 12070 + }, + { + "epoch": 0.37286252237792455, + "grad_norm": 0.28450822830200195, + "learning_rate": 9.386965403843168e-05, + "loss": 0.0558, + "step": 12080 + }, + { + "epoch": 0.3731711834063831, + "grad_norm": 0.27820804715156555, + "learning_rate": 9.385642589449726e-05, + "loss": 0.0548, + "step": 12090 + }, + { + "epoch": 0.37347984443484167, + "grad_norm": 0.49756893515586853, + "learning_rate": 9.38431844281386e-05, + "loss": 0.0591, + "step": 12100 + }, + { + "epoch": 0.3737885054633002, + "grad_norm": 0.3486950397491455, + "learning_rate": 9.38299296433781e-05, + "loss": 0.0537, + "step": 12110 + }, + { + "epoch": 0.3740971664917587, + "grad_norm": 0.30286362767219543, + "learning_rate": 9.381666154424226e-05, + "loss": 0.0563, + "step": 12120 + }, + { + "epoch": 0.3744058275202173, + "grad_norm": 0.35412469506263733, + "learning_rate": 9.380338013476157e-05, + "loss": 0.0556, + "step": 12130 + }, + { + "epoch": 0.37471448854867584, + "grad_norm": 0.2813165485858917, + "learning_rate": 9.379008541897054e-05, + "loss": 0.0546, + "step": 12140 + }, + { + "epoch": 0.3750231495771344, + "grad_norm": 0.5591005086898804, + "learning_rate": 9.377677740090777e-05, + "loss": 0.0558, + "step": 12150 + }, + { + "epoch": 0.37533181060559295, + "grad_norm": 0.2807675004005432, + "learning_rate": 9.376345608461588e-05, + "loss": 0.0575, + "step": 12160 + }, + { + "epoch": 0.3756404716340515, + "grad_norm": 0.3409714996814728, + "learning_rate": 9.375012147414155e-05, + "loss": 0.0562, + "step": 12170 + }, + { + "epoch": 0.37594913266251, + "grad_norm": 0.4050726592540741, + "learning_rate": 9.373677357353545e-05, + "loss": 0.0579, + "step": 12180 + }, + { + "epoch": 0.37625779369096857, + "grad_norm": 0.4824253022670746, + "learning_rate": 9.372341238685237e-05, + "loss": 0.0539, + "step": 12190 + }, + { + "epoch": 0.3765664547194271, + "grad_norm": 0.3462156653404236, + "learning_rate": 9.371003791815102e-05, + "loss": 0.0603, + "step": 12200 + }, + { + "epoch": 0.3768751157478857, + "grad_norm": 0.3427678048610687, + "learning_rate": 9.369665017149429e-05, + "loss": 0.0585, + "step": 12210 + }, + { + "epoch": 0.37718377677634424, + "grad_norm": 0.27239587903022766, + "learning_rate": 9.368324915094895e-05, + "loss": 0.051, + "step": 12220 + }, + { + "epoch": 0.37749243780480274, + "grad_norm": 0.3308522701263428, + "learning_rate": 9.366983486058591e-05, + "loss": 0.0547, + "step": 12230 + }, + { + "epoch": 0.3778010988332613, + "grad_norm": 0.3416873514652252, + "learning_rate": 9.365640730448009e-05, + "loss": 0.0585, + "step": 12240 + }, + { + "epoch": 0.37810975986171985, + "grad_norm": 0.3175276219844818, + "learning_rate": 9.36429664867104e-05, + "loss": 0.0507, + "step": 12250 + }, + { + "epoch": 0.3784184208901784, + "grad_norm": 0.26101478934288025, + "learning_rate": 9.362951241135982e-05, + "loss": 0.0553, + "step": 12260 + }, + { + "epoch": 0.37872708191863697, + "grad_norm": 0.3235292136669159, + "learning_rate": 9.361604508251534e-05, + "loss": 0.0622, + "step": 12270 + }, + { + "epoch": 0.3790357429470955, + "grad_norm": 0.3693544566631317, + "learning_rate": 9.360256450426799e-05, + "loss": 0.0545, + "step": 12280 + }, + { + "epoch": 0.379344403975554, + "grad_norm": 0.3163374960422516, + "learning_rate": 9.358907068071279e-05, + "loss": 0.0522, + "step": 12290 + }, + { + "epoch": 0.3796530650040126, + "grad_norm": 0.3828485906124115, + "learning_rate": 9.357556361594882e-05, + "loss": 0.0626, + "step": 12300 + }, + { + "epoch": 0.37996172603247114, + "grad_norm": 0.3111541271209717, + "learning_rate": 9.356204331407917e-05, + "loss": 0.0561, + "step": 12310 + }, + { + "epoch": 0.3802703870609297, + "grad_norm": 0.6581979990005493, + "learning_rate": 9.354850977921094e-05, + "loss": 0.0545, + "step": 12320 + }, + { + "epoch": 0.38057904808938825, + "grad_norm": 0.31079158186912537, + "learning_rate": 9.353496301545529e-05, + "loss": 0.051, + "step": 12330 + }, + { + "epoch": 0.3808877091178468, + "grad_norm": 0.25536078214645386, + "learning_rate": 9.352140302692733e-05, + "loss": 0.0581, + "step": 12340 + }, + { + "epoch": 0.3811963701463053, + "grad_norm": 0.43236178159713745, + "learning_rate": 9.350782981774627e-05, + "loss": 0.059, + "step": 12350 + }, + { + "epoch": 0.38150503117476386, + "grad_norm": 0.3646314740180969, + "learning_rate": 9.349424339203526e-05, + "loss": 0.0513, + "step": 12360 + }, + { + "epoch": 0.3818136922032224, + "grad_norm": 0.3358384966850281, + "learning_rate": 9.34806437539215e-05, + "loss": 0.0535, + "step": 12370 + }, + { + "epoch": 0.382122353231681, + "grad_norm": 0.5895518064498901, + "learning_rate": 9.346703090753622e-05, + "loss": 0.0525, + "step": 12380 + }, + { + "epoch": 0.38243101426013953, + "grad_norm": 0.2861016094684601, + "learning_rate": 9.345340485701461e-05, + "loss": 0.052, + "step": 12390 + }, + { + "epoch": 0.38273967528859804, + "grad_norm": 0.43261948227882385, + "learning_rate": 9.343976560649595e-05, + "loss": 0.0541, + "step": 12400 + }, + { + "epoch": 0.3830483363170566, + "grad_norm": 0.4083441495895386, + "learning_rate": 9.342611316012344e-05, + "loss": 0.0511, + "step": 12410 + }, + { + "epoch": 0.38335699734551515, + "grad_norm": 0.4615671634674072, + "learning_rate": 9.341244752204437e-05, + "loss": 0.0567, + "step": 12420 + }, + { + "epoch": 0.3836656583739737, + "grad_norm": 0.22125014662742615, + "learning_rate": 9.339876869640995e-05, + "loss": 0.0593, + "step": 12430 + }, + { + "epoch": 0.38397431940243226, + "grad_norm": 0.4114741086959839, + "learning_rate": 9.33850766873755e-05, + "loss": 0.0554, + "step": 12440 + }, + { + "epoch": 0.3842829804308908, + "grad_norm": 0.4117465019226074, + "learning_rate": 9.337137149910028e-05, + "loss": 0.0525, + "step": 12450 + }, + { + "epoch": 0.3845916414593493, + "grad_norm": 0.47657421231269836, + "learning_rate": 9.335765313574753e-05, + "loss": 0.0523, + "step": 12460 + }, + { + "epoch": 0.3849003024878079, + "grad_norm": 0.35144877433776855, + "learning_rate": 9.334392160148457e-05, + "loss": 0.0616, + "step": 12470 + }, + { + "epoch": 0.38520896351626643, + "grad_norm": 0.39570292830467224, + "learning_rate": 9.333017690048264e-05, + "loss": 0.0557, + "step": 12480 + }, + { + "epoch": 0.385517624544725, + "grad_norm": 0.29303091764450073, + "learning_rate": 9.331641903691706e-05, + "loss": 0.0534, + "step": 12490 + }, + { + "epoch": 0.38582628557318355, + "grad_norm": 0.23851604759693146, + "learning_rate": 9.330264801496707e-05, + "loss": 0.053, + "step": 12500 + }, + { + "epoch": 0.3861349466016421, + "grad_norm": 0.32327061891555786, + "learning_rate": 9.328886383881594e-05, + "loss": 0.0572, + "step": 12510 + }, + { + "epoch": 0.3864436076301006, + "grad_norm": 0.27697861194610596, + "learning_rate": 9.327506651265095e-05, + "loss": 0.0503, + "step": 12520 + }, + { + "epoch": 0.38675226865855916, + "grad_norm": 0.3737695813179016, + "learning_rate": 9.326125604066338e-05, + "loss": 0.057, + "step": 12530 + }, + { + "epoch": 0.3870609296870177, + "grad_norm": 0.3467184007167816, + "learning_rate": 9.324743242704847e-05, + "loss": 0.0536, + "step": 12540 + }, + { + "epoch": 0.3873695907154763, + "grad_norm": 0.32716843485832214, + "learning_rate": 9.323359567600546e-05, + "loss": 0.0523, + "step": 12550 + }, + { + "epoch": 0.38767825174393483, + "grad_norm": 1.7731115818023682, + "learning_rate": 9.321974579173761e-05, + "loss": 0.058, + "step": 12560 + }, + { + "epoch": 0.38798691277239333, + "grad_norm": 0.3070674240589142, + "learning_rate": 9.320588277845213e-05, + "loss": 0.0546, + "step": 12570 + }, + { + "epoch": 0.3882955738008519, + "grad_norm": 0.6498634219169617, + "learning_rate": 9.319200664036026e-05, + "loss": 0.0593, + "step": 12580 + }, + { + "epoch": 0.38860423482931045, + "grad_norm": 0.3568260967731476, + "learning_rate": 9.31781173816772e-05, + "loss": 0.0549, + "step": 12590 + }, + { + "epoch": 0.388912895857769, + "grad_norm": 0.4768182635307312, + "learning_rate": 9.316421500662212e-05, + "loss": 0.0549, + "step": 12600 + }, + { + "epoch": 0.38922155688622756, + "grad_norm": 0.43763673305511475, + "learning_rate": 9.31502995194182e-05, + "loss": 0.0501, + "step": 12610 + }, + { + "epoch": 0.3895302179146861, + "grad_norm": 0.35581961274147034, + "learning_rate": 9.31363709242926e-05, + "loss": 0.0535, + "step": 12620 + }, + { + "epoch": 0.3898388789431446, + "grad_norm": 0.4459213316440582, + "learning_rate": 9.312242922547647e-05, + "loss": 0.0556, + "step": 12630 + }, + { + "epoch": 0.3901475399716032, + "grad_norm": 0.3428115248680115, + "learning_rate": 9.310847442720492e-05, + "loss": 0.0585, + "step": 12640 + }, + { + "epoch": 0.39045620100006173, + "grad_norm": 0.3940505087375641, + "learning_rate": 9.309450653371706e-05, + "loss": 0.0537, + "step": 12650 + }, + { + "epoch": 0.3907648620285203, + "grad_norm": 0.3694036304950714, + "learning_rate": 9.308052554925595e-05, + "loss": 0.0544, + "step": 12660 + }, + { + "epoch": 0.39107352305697884, + "grad_norm": 0.3235882520675659, + "learning_rate": 9.306653147806867e-05, + "loss": 0.055, + "step": 12670 + }, + { + "epoch": 0.39138218408543735, + "grad_norm": 0.32800450921058655, + "learning_rate": 9.305252432440622e-05, + "loss": 0.0549, + "step": 12680 + }, + { + "epoch": 0.3916908451138959, + "grad_norm": 0.24939344823360443, + "learning_rate": 9.303850409252361e-05, + "loss": 0.0592, + "step": 12690 + }, + { + "epoch": 0.39199950614235446, + "grad_norm": 0.284182608127594, + "learning_rate": 9.302447078667985e-05, + "loss": 0.0512, + "step": 12700 + }, + { + "epoch": 0.392308167170813, + "grad_norm": 0.2993795573711395, + "learning_rate": 9.301042441113783e-05, + "loss": 0.0527, + "step": 12710 + }, + { + "epoch": 0.3926168281992716, + "grad_norm": 0.36127546429634094, + "learning_rate": 9.299636497016451e-05, + "loss": 0.0562, + "step": 12720 + }, + { + "epoch": 0.39292548922773013, + "grad_norm": 0.48526063561439514, + "learning_rate": 9.298229246803076e-05, + "loss": 0.0567, + "step": 12730 + }, + { + "epoch": 0.39323415025618863, + "grad_norm": 0.6731045842170715, + "learning_rate": 9.296820690901144e-05, + "loss": 0.052, + "step": 12740 + }, + { + "epoch": 0.3935428112846472, + "grad_norm": 0.44754183292388916, + "learning_rate": 9.295410829738539e-05, + "loss": 0.056, + "step": 12750 + }, + { + "epoch": 0.39385147231310574, + "grad_norm": 0.33990979194641113, + "learning_rate": 9.293999663743535e-05, + "loss": 0.06, + "step": 12760 + }, + { + "epoch": 0.3941601333415643, + "grad_norm": 0.3213772177696228, + "learning_rate": 9.292587193344813e-05, + "loss": 0.0604, + "step": 12770 + }, + { + "epoch": 0.39446879437002286, + "grad_norm": 0.2711023986339569, + "learning_rate": 9.291173418971437e-05, + "loss": 0.0547, + "step": 12780 + }, + { + "epoch": 0.3947774553984814, + "grad_norm": 0.3094342052936554, + "learning_rate": 9.28975834105288e-05, + "loss": 0.0588, + "step": 12790 + }, + { + "epoch": 0.3950861164269399, + "grad_norm": 0.34757816791534424, + "learning_rate": 9.288341960019004e-05, + "loss": 0.0556, + "step": 12800 + }, + { + "epoch": 0.39539477745539847, + "grad_norm": 0.3270135223865509, + "learning_rate": 9.286924276300067e-05, + "loss": 0.0544, + "step": 12810 + }, + { + "epoch": 0.39570343848385703, + "grad_norm": 0.26874884963035583, + "learning_rate": 9.285505290326726e-05, + "loss": 0.0546, + "step": 12820 + }, + { + "epoch": 0.3960120995123156, + "grad_norm": 0.2646600902080536, + "learning_rate": 9.284085002530027e-05, + "loss": 0.05, + "step": 12830 + }, + { + "epoch": 0.39632076054077414, + "grad_norm": 0.2560214698314667, + "learning_rate": 9.282663413341422e-05, + "loss": 0.0522, + "step": 12840 + }, + { + "epoch": 0.39662942156923264, + "grad_norm": 0.3405613601207733, + "learning_rate": 9.281240523192747e-05, + "loss": 0.059, + "step": 12850 + }, + { + "epoch": 0.3969380825976912, + "grad_norm": 0.3459934592247009, + "learning_rate": 9.279816332516242e-05, + "loss": 0.0563, + "step": 12860 + }, + { + "epoch": 0.39724674362614976, + "grad_norm": 0.34183570742607117, + "learning_rate": 9.278390841744536e-05, + "loss": 0.0542, + "step": 12870 + }, + { + "epoch": 0.3975554046546083, + "grad_norm": 0.4106989800930023, + "learning_rate": 9.276964051310658e-05, + "loss": 0.0515, + "step": 12880 + }, + { + "epoch": 0.39786406568306687, + "grad_norm": 0.31743499636650085, + "learning_rate": 9.275535961648027e-05, + "loss": 0.058, + "step": 12890 + }, + { + "epoch": 0.3981727267115254, + "grad_norm": 0.5247482061386108, + "learning_rate": 9.274106573190459e-05, + "loss": 0.0535, + "step": 12900 + }, + { + "epoch": 0.3984813877399839, + "grad_norm": 0.29644671082496643, + "learning_rate": 9.272675886372168e-05, + "loss": 0.0523, + "step": 12910 + }, + { + "epoch": 0.3987900487684425, + "grad_norm": 0.48666688799858093, + "learning_rate": 9.271243901627754e-05, + "loss": 0.0565, + "step": 12920 + }, + { + "epoch": 0.39909870979690104, + "grad_norm": 0.33581429719924927, + "learning_rate": 9.269810619392219e-05, + "loss": 0.0619, + "step": 12930 + }, + { + "epoch": 0.3994073708253596, + "grad_norm": 0.4892466068267822, + "learning_rate": 9.268376040100955e-05, + "loss": 0.0556, + "step": 12940 + }, + { + "epoch": 0.39971603185381815, + "grad_norm": 0.260172575712204, + "learning_rate": 9.266940164189752e-05, + "loss": 0.0568, + "step": 12950 + }, + { + "epoch": 0.4000246928822767, + "grad_norm": 0.4344863295555115, + "learning_rate": 9.265502992094787e-05, + "loss": 0.0531, + "step": 12960 + }, + { + "epoch": 0.4003333539107352, + "grad_norm": 0.2985997200012207, + "learning_rate": 9.264064524252638e-05, + "loss": 0.0555, + "step": 12970 + }, + { + "epoch": 0.40064201493919377, + "grad_norm": 0.29818663001060486, + "learning_rate": 9.262624761100271e-05, + "loss": 0.0543, + "step": 12980 + }, + { + "epoch": 0.4009506759676523, + "grad_norm": 0.33637672662734985, + "learning_rate": 9.261183703075051e-05, + "loss": 0.0539, + "step": 12990 + }, + { + "epoch": 0.4012593369961109, + "grad_norm": 0.333720326423645, + "learning_rate": 9.259741350614733e-05, + "loss": 0.0546, + "step": 13000 + }, + { + "epoch": 0.40156799802456944, + "grad_norm": 0.36403271555900574, + "learning_rate": 9.258297704157464e-05, + "loss": 0.0541, + "step": 13010 + }, + { + "epoch": 0.40187665905302794, + "grad_norm": 0.38550427556037903, + "learning_rate": 9.256852764141786e-05, + "loss": 0.0584, + "step": 13020 + }, + { + "epoch": 0.4021853200814865, + "grad_norm": 0.3865271508693695, + "learning_rate": 9.255406531006634e-05, + "loss": 0.0518, + "step": 13030 + }, + { + "epoch": 0.40249398110994505, + "grad_norm": 0.33559948205947876, + "learning_rate": 9.253959005191335e-05, + "loss": 0.0527, + "step": 13040 + }, + { + "epoch": 0.4028026421384036, + "grad_norm": 0.27854475378990173, + "learning_rate": 9.25251018713561e-05, + "loss": 0.0565, + "step": 13050 + }, + { + "epoch": 0.40311130316686217, + "grad_norm": 0.393358439207077, + "learning_rate": 9.251060077279571e-05, + "loss": 0.0568, + "step": 13060 + }, + { + "epoch": 0.4034199641953207, + "grad_norm": 0.37384408712387085, + "learning_rate": 9.249608676063724e-05, + "loss": 0.0476, + "step": 13070 + }, + { + "epoch": 0.4037286252237792, + "grad_norm": 0.4203333854675293, + "learning_rate": 9.248155983928964e-05, + "loss": 0.0594, + "step": 13080 + }, + { + "epoch": 0.4040372862522378, + "grad_norm": 0.3548599183559418, + "learning_rate": 9.246702001316583e-05, + "loss": 0.0562, + "step": 13090 + }, + { + "epoch": 0.40434594728069634, + "grad_norm": 0.3006739616394043, + "learning_rate": 9.245246728668262e-05, + "loss": 0.0557, + "step": 13100 + }, + { + "epoch": 0.4046546083091549, + "grad_norm": 0.3816075026988983, + "learning_rate": 9.243790166426073e-05, + "loss": 0.0565, + "step": 13110 + }, + { + "epoch": 0.40496326933761345, + "grad_norm": 0.2843579053878784, + "learning_rate": 9.242332315032484e-05, + "loss": 0.0542, + "step": 13120 + }, + { + "epoch": 0.40527193036607195, + "grad_norm": 0.49494418501853943, + "learning_rate": 9.240873174930349e-05, + "loss": 0.0556, + "step": 13130 + }, + { + "epoch": 0.4055805913945305, + "grad_norm": 0.31489303708076477, + "learning_rate": 9.239412746562917e-05, + "loss": 0.049, + "step": 13140 + }, + { + "epoch": 0.40588925242298907, + "grad_norm": 0.30179324746131897, + "learning_rate": 9.237951030373828e-05, + "loss": 0.0523, + "step": 13150 + }, + { + "epoch": 0.4061979134514476, + "grad_norm": 0.3610905110836029, + "learning_rate": 9.236488026807113e-05, + "loss": 0.051, + "step": 13160 + }, + { + "epoch": 0.4065065744799062, + "grad_norm": 0.45784521102905273, + "learning_rate": 9.235023736307193e-05, + "loss": 0.0586, + "step": 13170 + }, + { + "epoch": 0.40681523550836474, + "grad_norm": 0.3227314352989197, + "learning_rate": 9.233558159318881e-05, + "loss": 0.0564, + "step": 13180 + }, + { + "epoch": 0.40712389653682324, + "grad_norm": 0.30500325560569763, + "learning_rate": 9.232091296287382e-05, + "loss": 0.0576, + "step": 13190 + }, + { + "epoch": 0.4074325575652818, + "grad_norm": 0.29457610845565796, + "learning_rate": 9.230623147658288e-05, + "loss": 0.0476, + "step": 13200 + }, + { + "epoch": 0.40774121859374035, + "grad_norm": 0.3227302134037018, + "learning_rate": 9.229153713877586e-05, + "loss": 0.0531, + "step": 13210 + }, + { + "epoch": 0.4080498796221989, + "grad_norm": 0.3533036410808563, + "learning_rate": 9.227682995391649e-05, + "loss": 0.0546, + "step": 13220 + }, + { + "epoch": 0.40835854065065746, + "grad_norm": 0.26626354455947876, + "learning_rate": 9.226210992647243e-05, + "loss": 0.0562, + "step": 13230 + }, + { + "epoch": 0.408667201679116, + "grad_norm": 0.4204142093658447, + "learning_rate": 9.224737706091525e-05, + "loss": 0.0522, + "step": 13240 + }, + { + "epoch": 0.4089758627075745, + "grad_norm": 0.3012014627456665, + "learning_rate": 9.223263136172039e-05, + "loss": 0.0546, + "step": 13250 + }, + { + "epoch": 0.4092845237360331, + "grad_norm": 0.36932870745658875, + "learning_rate": 9.22178728333672e-05, + "loss": 0.058, + "step": 13260 + }, + { + "epoch": 0.40959318476449164, + "grad_norm": 0.3215840756893158, + "learning_rate": 9.220310148033897e-05, + "loss": 0.0507, + "step": 13270 + }, + { + "epoch": 0.4099018457929502, + "grad_norm": 0.25983062386512756, + "learning_rate": 9.21883173071228e-05, + "loss": 0.058, + "step": 13280 + }, + { + "epoch": 0.41021050682140875, + "grad_norm": 0.45934924483299255, + "learning_rate": 9.217352031820976e-05, + "loss": 0.0583, + "step": 13290 + }, + { + "epoch": 0.41051916784986725, + "grad_norm": 0.2710789442062378, + "learning_rate": 9.215871051809477e-05, + "loss": 0.0545, + "step": 13300 + }, + { + "epoch": 0.4108278288783258, + "grad_norm": 0.3719940483570099, + "learning_rate": 9.214388791127666e-05, + "loss": 0.0531, + "step": 13310 + }, + { + "epoch": 0.41113648990678436, + "grad_norm": 0.3286314904689789, + "learning_rate": 9.212905250225814e-05, + "loss": 0.0605, + "step": 13320 + }, + { + "epoch": 0.4114451509352429, + "grad_norm": 0.4469938576221466, + "learning_rate": 9.211420429554583e-05, + "loss": 0.0575, + "step": 13330 + }, + { + "epoch": 0.4117538119637015, + "grad_norm": 0.31993186473846436, + "learning_rate": 9.209934329565022e-05, + "loss": 0.0536, + "step": 13340 + }, + { + "epoch": 0.41206247299216003, + "grad_norm": 0.253307044506073, + "learning_rate": 9.208446950708568e-05, + "loss": 0.054, + "step": 13350 + }, + { + "epoch": 0.41237113402061853, + "grad_norm": 0.4968326985836029, + "learning_rate": 9.20695829343705e-05, + "loss": 0.0616, + "step": 13360 + }, + { + "epoch": 0.4126797950490771, + "grad_norm": 0.3253658413887024, + "learning_rate": 9.205468358202678e-05, + "loss": 0.0521, + "step": 13370 + }, + { + "epoch": 0.41298845607753565, + "grad_norm": 0.3072507083415985, + "learning_rate": 9.203977145458059e-05, + "loss": 0.0535, + "step": 13380 + }, + { + "epoch": 0.4132971171059942, + "grad_norm": 0.34612563252449036, + "learning_rate": 9.202484655656182e-05, + "loss": 0.0564, + "step": 13390 + }, + { + "epoch": 0.41360577813445276, + "grad_norm": 0.3739696741104126, + "learning_rate": 9.200990889250427e-05, + "loss": 0.0567, + "step": 13400 + }, + { + "epoch": 0.4139144391629113, + "grad_norm": 0.2707541286945343, + "learning_rate": 9.19949584669456e-05, + "loss": 0.0527, + "step": 13410 + }, + { + "epoch": 0.4142231001913698, + "grad_norm": 0.29661595821380615, + "learning_rate": 9.197999528442738e-05, + "loss": 0.0564, + "step": 13420 + }, + { + "epoch": 0.4145317612198284, + "grad_norm": 0.3214598000049591, + "learning_rate": 9.196501934949499e-05, + "loss": 0.0582, + "step": 13430 + }, + { + "epoch": 0.41484042224828693, + "grad_norm": 0.3288394510746002, + "learning_rate": 9.195003066669776e-05, + "loss": 0.0568, + "step": 13440 + }, + { + "epoch": 0.4151490832767455, + "grad_norm": 0.30615949630737305, + "learning_rate": 9.193502924058884e-05, + "loss": 0.0505, + "step": 13450 + }, + { + "epoch": 0.41545774430520405, + "grad_norm": 0.3628118634223938, + "learning_rate": 9.192001507572526e-05, + "loss": 0.0518, + "step": 13460 + }, + { + "epoch": 0.41576640533366255, + "grad_norm": 0.40468138456344604, + "learning_rate": 9.190498817666793e-05, + "loss": 0.0547, + "step": 13470 + }, + { + "epoch": 0.4160750663621211, + "grad_norm": 0.28810006380081177, + "learning_rate": 9.188994854798163e-05, + "loss": 0.0589, + "step": 13480 + }, + { + "epoch": 0.41638372739057966, + "grad_norm": 0.31277215480804443, + "learning_rate": 9.187489619423499e-05, + "loss": 0.0514, + "step": 13490 + }, + { + "epoch": 0.4166923884190382, + "grad_norm": 0.3146766424179077, + "learning_rate": 9.185983112000056e-05, + "loss": 0.058, + "step": 13500 + }, + { + "epoch": 0.4170010494474968, + "grad_norm": 0.27181750535964966, + "learning_rate": 9.184475332985464e-05, + "loss": 0.0574, + "step": 13510 + }, + { + "epoch": 0.41730971047595533, + "grad_norm": 0.3045525550842285, + "learning_rate": 9.182966282837754e-05, + "loss": 0.0519, + "step": 13520 + }, + { + "epoch": 0.41761837150441383, + "grad_norm": 0.4577593207359314, + "learning_rate": 9.18145596201533e-05, + "loss": 0.0533, + "step": 13530 + }, + { + "epoch": 0.4179270325328724, + "grad_norm": 0.35696083307266235, + "learning_rate": 9.179944370976991e-05, + "loss": 0.0554, + "step": 13540 + }, + { + "epoch": 0.41823569356133095, + "grad_norm": 0.2647937536239624, + "learning_rate": 9.178431510181918e-05, + "loss": 0.0546, + "step": 13550 + }, + { + "epoch": 0.4185443545897895, + "grad_norm": 0.28393658995628357, + "learning_rate": 9.176917380089675e-05, + "loss": 0.055, + "step": 13560 + }, + { + "epoch": 0.41885301561824806, + "grad_norm": 0.4663776159286499, + "learning_rate": 9.175401981160219e-05, + "loss": 0.0575, + "step": 13570 + }, + { + "epoch": 0.41916167664670656, + "grad_norm": 0.34712663292884827, + "learning_rate": 9.173885313853885e-05, + "loss": 0.0559, + "step": 13580 + }, + { + "epoch": 0.4194703376751651, + "grad_norm": 0.29325613379478455, + "learning_rate": 9.172367378631398e-05, + "loss": 0.056, + "step": 13590 + }, + { + "epoch": 0.4197789987036237, + "grad_norm": 0.5325918793678284, + "learning_rate": 9.170848175953866e-05, + "loss": 0.0576, + "step": 13600 + }, + { + "epoch": 0.42008765973208223, + "grad_norm": 0.3100646436214447, + "learning_rate": 9.169327706282784e-05, + "loss": 0.0509, + "step": 13610 + }, + { + "epoch": 0.4203963207605408, + "grad_norm": 0.3728489577770233, + "learning_rate": 9.167805970080029e-05, + "loss": 0.0525, + "step": 13620 + }, + { + "epoch": 0.42070498178899934, + "grad_norm": 0.350833922624588, + "learning_rate": 9.166282967807864e-05, + "loss": 0.0527, + "step": 13630 + }, + { + "epoch": 0.42101364281745784, + "grad_norm": 0.3330201506614685, + "learning_rate": 9.16475869992894e-05, + "loss": 0.054, + "step": 13640 + }, + { + "epoch": 0.4213223038459164, + "grad_norm": 0.5681411623954773, + "learning_rate": 9.163233166906284e-05, + "loss": 0.059, + "step": 13650 + }, + { + "epoch": 0.42163096487437496, + "grad_norm": 0.4063504636287689, + "learning_rate": 9.161706369203317e-05, + "loss": 0.0531, + "step": 13660 + }, + { + "epoch": 0.4219396259028335, + "grad_norm": 0.3135746121406555, + "learning_rate": 9.16017830728384e-05, + "loss": 0.0532, + "step": 13670 + }, + { + "epoch": 0.42224828693129207, + "grad_norm": 0.34629005193710327, + "learning_rate": 9.158648981612035e-05, + "loss": 0.0616, + "step": 13680 + }, + { + "epoch": 0.42255694795975063, + "grad_norm": 0.32041165232658386, + "learning_rate": 9.157118392652472e-05, + "loss": 0.0577, + "step": 13690 + }, + { + "epoch": 0.42286560898820913, + "grad_norm": 0.39504626393318176, + "learning_rate": 9.155586540870104e-05, + "loss": 0.0573, + "step": 13700 + }, + { + "epoch": 0.4231742700166677, + "grad_norm": 0.3467038571834564, + "learning_rate": 9.154053426730267e-05, + "loss": 0.0573, + "step": 13710 + }, + { + "epoch": 0.42348293104512624, + "grad_norm": 0.424021452665329, + "learning_rate": 9.15251905069868e-05, + "loss": 0.0487, + "step": 13720 + }, + { + "epoch": 0.4237915920735848, + "grad_norm": 0.4883821904659271, + "learning_rate": 9.150983413241446e-05, + "loss": 0.0569, + "step": 13730 + }, + { + "epoch": 0.42410025310204336, + "grad_norm": 0.29530853033065796, + "learning_rate": 9.149446514825051e-05, + "loss": 0.0534, + "step": 13740 + }, + { + "epoch": 0.42440891413050186, + "grad_norm": 0.5646997690200806, + "learning_rate": 9.147908355916365e-05, + "loss": 0.0557, + "step": 13750 + }, + { + "epoch": 0.4247175751589604, + "grad_norm": 0.6703916192054749, + "learning_rate": 9.146368936982642e-05, + "loss": 0.0551, + "step": 13760 + }, + { + "epoch": 0.42502623618741897, + "grad_norm": 0.3190588355064392, + "learning_rate": 9.144828258491511e-05, + "loss": 0.0517, + "step": 13770 + }, + { + "epoch": 0.4253348972158775, + "grad_norm": 0.4708987772464752, + "learning_rate": 9.143286320910996e-05, + "loss": 0.051, + "step": 13780 + }, + { + "epoch": 0.4256435582443361, + "grad_norm": 0.27706968784332275, + "learning_rate": 9.141743124709491e-05, + "loss": 0.053, + "step": 13790 + }, + { + "epoch": 0.42595221927279464, + "grad_norm": 0.25759851932525635, + "learning_rate": 9.140198670355784e-05, + "loss": 0.0537, + "step": 13800 + }, + { + "epoch": 0.42626088030125314, + "grad_norm": 0.3362017273902893, + "learning_rate": 9.138652958319034e-05, + "loss": 0.0544, + "step": 13810 + }, + { + "epoch": 0.4265695413297117, + "grad_norm": 0.3387603759765625, + "learning_rate": 9.137105989068791e-05, + "loss": 0.0545, + "step": 13820 + }, + { + "epoch": 0.42687820235817026, + "grad_norm": 0.33943361043930054, + "learning_rate": 9.135557763074983e-05, + "loss": 0.0569, + "step": 13830 + }, + { + "epoch": 0.4271868633866288, + "grad_norm": 0.4045940041542053, + "learning_rate": 9.13400828080792e-05, + "loss": 0.0575, + "step": 13840 + }, + { + "epoch": 0.42749552441508737, + "grad_norm": 0.3776386082172394, + "learning_rate": 9.132457542738292e-05, + "loss": 0.051, + "step": 13850 + }, + { + "epoch": 0.42780418544354587, + "grad_norm": 0.29894113540649414, + "learning_rate": 9.130905549337174e-05, + "loss": 0.0491, + "step": 13860 + }, + { + "epoch": 0.4281128464720044, + "grad_norm": 0.3044470250606537, + "learning_rate": 9.129352301076021e-05, + "loss": 0.0566, + "step": 13870 + }, + { + "epoch": 0.428421507500463, + "grad_norm": 0.23303329944610596, + "learning_rate": 9.127797798426668e-05, + "loss": 0.0547, + "step": 13880 + }, + { + "epoch": 0.42873016852892154, + "grad_norm": 0.48370951414108276, + "learning_rate": 9.126242041861333e-05, + "loss": 0.0543, + "step": 13890 + }, + { + "epoch": 0.4290388295573801, + "grad_norm": 0.2670044004917145, + "learning_rate": 9.124685031852611e-05, + "loss": 0.0494, + "step": 13900 + }, + { + "epoch": 0.42934749058583865, + "grad_norm": 0.3032512962818146, + "learning_rate": 9.123126768873482e-05, + "loss": 0.0602, + "step": 13910 + }, + { + "epoch": 0.42965615161429715, + "grad_norm": 0.44255802035331726, + "learning_rate": 9.121567253397308e-05, + "loss": 0.0525, + "step": 13920 + }, + { + "epoch": 0.4299648126427557, + "grad_norm": 0.3213754892349243, + "learning_rate": 9.120006485897824e-05, + "loss": 0.0499, + "step": 13930 + }, + { + "epoch": 0.43027347367121427, + "grad_norm": 0.35068652033805847, + "learning_rate": 9.118444466849152e-05, + "loss": 0.051, + "step": 13940 + }, + { + "epoch": 0.4305821346996728, + "grad_norm": 0.37582212686538696, + "learning_rate": 9.116881196725793e-05, + "loss": 0.0537, + "step": 13950 + }, + { + "epoch": 0.4308907957281314, + "grad_norm": 0.38775497674942017, + "learning_rate": 9.115316676002627e-05, + "loss": 0.0557, + "step": 13960 + }, + { + "epoch": 0.43119945675658994, + "grad_norm": 0.38999414443969727, + "learning_rate": 9.113750905154911e-05, + "loss": 0.0568, + "step": 13970 + }, + { + "epoch": 0.43150811778504844, + "grad_norm": 0.38092732429504395, + "learning_rate": 9.112183884658289e-05, + "loss": 0.0503, + "step": 13980 + }, + { + "epoch": 0.431816778813507, + "grad_norm": 0.37590259313583374, + "learning_rate": 9.11061561498878e-05, + "loss": 0.056, + "step": 13990 + }, + { + "epoch": 0.43212543984196555, + "grad_norm": 0.30347782373428345, + "learning_rate": 9.109046096622779e-05, + "loss": 0.0493, + "step": 14000 + }, + { + "epoch": 0.4324341008704241, + "grad_norm": 0.34489649534225464, + "learning_rate": 9.107475330037069e-05, + "loss": 0.056, + "step": 14010 + }, + { + "epoch": 0.43274276189888267, + "grad_norm": 0.5815978050231934, + "learning_rate": 9.105903315708806e-05, + "loss": 0.0552, + "step": 14020 + }, + { + "epoch": 0.43305142292734117, + "grad_norm": 0.27338266372680664, + "learning_rate": 9.104330054115524e-05, + "loss": 0.0552, + "step": 14030 + }, + { + "epoch": 0.4333600839557997, + "grad_norm": 0.36796092987060547, + "learning_rate": 9.102755545735141e-05, + "loss": 0.0506, + "step": 14040 + }, + { + "epoch": 0.4336687449842583, + "grad_norm": 0.29348325729370117, + "learning_rate": 9.10117979104595e-05, + "loss": 0.0584, + "step": 14050 + }, + { + "epoch": 0.43397740601271684, + "grad_norm": 0.34248483180999756, + "learning_rate": 9.099602790526624e-05, + "loss": 0.0539, + "step": 14060 + }, + { + "epoch": 0.4342860670411754, + "grad_norm": 0.2721204459667206, + "learning_rate": 9.098024544656212e-05, + "loss": 0.0478, + "step": 14070 + }, + { + "epoch": 0.43459472806963395, + "grad_norm": 0.3065623939037323, + "learning_rate": 9.096445053914148e-05, + "loss": 0.0492, + "step": 14080 + }, + { + "epoch": 0.43490338909809245, + "grad_norm": 0.3335457146167755, + "learning_rate": 9.094864318780236e-05, + "loss": 0.0539, + "step": 14090 + }, + { + "epoch": 0.435212050126551, + "grad_norm": 0.26013603806495667, + "learning_rate": 9.093282339734663e-05, + "loss": 0.0515, + "step": 14100 + }, + { + "epoch": 0.43552071115500957, + "grad_norm": 0.3797277808189392, + "learning_rate": 9.091699117257992e-05, + "loss": 0.0563, + "step": 14110 + }, + { + "epoch": 0.4358293721834681, + "grad_norm": 0.3186219334602356, + "learning_rate": 9.090114651831163e-05, + "loss": 0.054, + "step": 14120 + }, + { + "epoch": 0.4361380332119267, + "grad_norm": 0.28969690203666687, + "learning_rate": 9.088528943935497e-05, + "loss": 0.0602, + "step": 14130 + }, + { + "epoch": 0.43644669424038524, + "grad_norm": 0.31149736046791077, + "learning_rate": 9.086941994052689e-05, + "loss": 0.0527, + "step": 14140 + }, + { + "epoch": 0.43675535526884374, + "grad_norm": 0.31453821063041687, + "learning_rate": 9.085353802664813e-05, + "loss": 0.05, + "step": 14150 + }, + { + "epoch": 0.4370640162973023, + "grad_norm": 0.2760377824306488, + "learning_rate": 9.08376437025432e-05, + "loss": 0.0525, + "step": 14160 + }, + { + "epoch": 0.43737267732576085, + "grad_norm": 0.3533741533756256, + "learning_rate": 9.082173697304035e-05, + "loss": 0.0538, + "step": 14170 + }, + { + "epoch": 0.4376813383542194, + "grad_norm": 0.29783859848976135, + "learning_rate": 9.080581784297166e-05, + "loss": 0.0525, + "step": 14180 + }, + { + "epoch": 0.43798999938267796, + "grad_norm": 0.3363410532474518, + "learning_rate": 9.078988631717291e-05, + "loss": 0.0508, + "step": 14190 + }, + { + "epoch": 0.43829866041113646, + "grad_norm": 0.3350031077861786, + "learning_rate": 9.077394240048369e-05, + "loss": 0.0496, + "step": 14200 + }, + { + "epoch": 0.438607321439595, + "grad_norm": 0.2860619127750397, + "learning_rate": 9.075798609774736e-05, + "loss": 0.0535, + "step": 14210 + }, + { + "epoch": 0.4389159824680536, + "grad_norm": 0.32524827122688293, + "learning_rate": 9.0742017413811e-05, + "loss": 0.0541, + "step": 14220 + }, + { + "epoch": 0.43922464349651213, + "grad_norm": 0.3235805034637451, + "learning_rate": 9.072603635352548e-05, + "loss": 0.0561, + "step": 14230 + }, + { + "epoch": 0.4395333045249707, + "grad_norm": 0.41083118319511414, + "learning_rate": 9.071004292174541e-05, + "loss": 0.0533, + "step": 14240 + }, + { + "epoch": 0.43984196555342925, + "grad_norm": 0.29432955384254456, + "learning_rate": 9.06940371233292e-05, + "loss": 0.0568, + "step": 14250 + }, + { + "epoch": 0.44015062658188775, + "grad_norm": 0.35038670897483826, + "learning_rate": 9.067801896313898e-05, + "loss": 0.0531, + "step": 14260 + }, + { + "epoch": 0.4404592876103463, + "grad_norm": 0.3495105803012848, + "learning_rate": 9.066198844604064e-05, + "loss": 0.0541, + "step": 14270 + }, + { + "epoch": 0.44076794863880486, + "grad_norm": 0.31105369329452515, + "learning_rate": 9.06459455769038e-05, + "loss": 0.0509, + "step": 14280 + }, + { + "epoch": 0.4410766096672634, + "grad_norm": 0.26353225111961365, + "learning_rate": 9.062989036060193e-05, + "loss": 0.0527, + "step": 14290 + }, + { + "epoch": 0.441385270695722, + "grad_norm": 0.40967896580696106, + "learning_rate": 9.061382280201212e-05, + "loss": 0.0508, + "step": 14300 + }, + { + "epoch": 0.4416939317241805, + "grad_norm": 0.369657963514328, + "learning_rate": 9.059774290601528e-05, + "loss": 0.0588, + "step": 14310 + }, + { + "epoch": 0.44200259275263903, + "grad_norm": 0.2999798655509949, + "learning_rate": 9.058165067749606e-05, + "loss": 0.055, + "step": 14320 + }, + { + "epoch": 0.4423112537810976, + "grad_norm": 0.268953412771225, + "learning_rate": 9.056554612134288e-05, + "loss": 0.0503, + "step": 14330 + }, + { + "epoch": 0.44261991480955615, + "grad_norm": 0.4222799241542816, + "learning_rate": 9.054942924244785e-05, + "loss": 0.0564, + "step": 14340 + }, + { + "epoch": 0.4429285758380147, + "grad_norm": 0.2386035919189453, + "learning_rate": 9.053330004570686e-05, + "loss": 0.0518, + "step": 14350 + }, + { + "epoch": 0.44323723686647326, + "grad_norm": 0.3262602984905243, + "learning_rate": 9.051715853601955e-05, + "loss": 0.0538, + "step": 14360 + }, + { + "epoch": 0.44354589789493176, + "grad_norm": 0.30174967646598816, + "learning_rate": 9.050100471828926e-05, + "loss": 0.0513, + "step": 14370 + }, + { + "epoch": 0.4438545589233903, + "grad_norm": 0.41442254185676575, + "learning_rate": 9.048483859742311e-05, + "loss": 0.0562, + "step": 14380 + }, + { + "epoch": 0.4441632199518489, + "grad_norm": 0.3615477383136749, + "learning_rate": 9.046866017833193e-05, + "loss": 0.0571, + "step": 14390 + }, + { + "epoch": 0.44447188098030743, + "grad_norm": 0.2846594452857971, + "learning_rate": 9.045246946593029e-05, + "loss": 0.0532, + "step": 14400 + }, + { + "epoch": 0.444780542008766, + "grad_norm": 0.29935315251350403, + "learning_rate": 9.043626646513652e-05, + "loss": 0.0532, + "step": 14410 + }, + { + "epoch": 0.44508920303722455, + "grad_norm": 0.32359808683395386, + "learning_rate": 9.042005118087267e-05, + "loss": 0.0489, + "step": 14420 + }, + { + "epoch": 0.44539786406568305, + "grad_norm": 0.5630438923835754, + "learning_rate": 9.040382361806448e-05, + "loss": 0.0553, + "step": 14430 + }, + { + "epoch": 0.4457065250941416, + "grad_norm": 0.32072892785072327, + "learning_rate": 9.038758378164148e-05, + "loss": 0.0573, + "step": 14440 + }, + { + "epoch": 0.44601518612260016, + "grad_norm": 0.35980749130249023, + "learning_rate": 9.037133167653691e-05, + "loss": 0.0551, + "step": 14450 + }, + { + "epoch": 0.4463238471510587, + "grad_norm": 0.3329438269138336, + "learning_rate": 9.035506730768771e-05, + "loss": 0.0543, + "step": 14460 + }, + { + "epoch": 0.4466325081795173, + "grad_norm": 0.31610164046287537, + "learning_rate": 9.033879068003458e-05, + "loss": 0.0515, + "step": 14470 + }, + { + "epoch": 0.4469411692079758, + "grad_norm": 0.41997915506362915, + "learning_rate": 9.032250179852193e-05, + "loss": 0.0506, + "step": 14480 + }, + { + "epoch": 0.44724983023643433, + "grad_norm": 0.44062918424606323, + "learning_rate": 9.030620066809787e-05, + "loss": 0.0552, + "step": 14490 + }, + { + "epoch": 0.4475584912648929, + "grad_norm": 0.28029748797416687, + "learning_rate": 9.028988729371428e-05, + "loss": 0.056, + "step": 14500 + }, + { + "epoch": 0.44786715229335144, + "grad_norm": 0.29415836930274963, + "learning_rate": 9.027356168032673e-05, + "loss": 0.0562, + "step": 14510 + }, + { + "epoch": 0.44817581332181, + "grad_norm": 0.30766239762306213, + "learning_rate": 9.02572238328945e-05, + "loss": 0.0527, + "step": 14520 + }, + { + "epoch": 0.44848447435026856, + "grad_norm": 0.45137178897857666, + "learning_rate": 9.02408737563806e-05, + "loss": 0.0542, + "step": 14530 + }, + { + "epoch": 0.44879313537872706, + "grad_norm": 0.32772669196128845, + "learning_rate": 9.022451145575174e-05, + "loss": 0.0509, + "step": 14540 + }, + { + "epoch": 0.4491017964071856, + "grad_norm": 0.25239458680152893, + "learning_rate": 9.02081369359784e-05, + "loss": 0.0615, + "step": 14550 + }, + { + "epoch": 0.4494104574356442, + "grad_norm": 0.3518636226654053, + "learning_rate": 9.019175020203465e-05, + "loss": 0.0558, + "step": 14560 + }, + { + "epoch": 0.44971911846410273, + "grad_norm": 0.333443820476532, + "learning_rate": 9.017535125889842e-05, + "loss": 0.0543, + "step": 14570 + }, + { + "epoch": 0.4500277794925613, + "grad_norm": 0.2723519504070282, + "learning_rate": 9.015894011155124e-05, + "loss": 0.0502, + "step": 14580 + }, + { + "epoch": 0.45033644052101984, + "grad_norm": 0.2840729057788849, + "learning_rate": 9.014251676497838e-05, + "loss": 0.0556, + "step": 14590 + }, + { + "epoch": 0.45064510154947834, + "grad_norm": 0.3935900926589966, + "learning_rate": 9.012608122416884e-05, + "loss": 0.057, + "step": 14600 + }, + { + "epoch": 0.4509537625779369, + "grad_norm": 0.6857351064682007, + "learning_rate": 9.010963349411529e-05, + "loss": 0.0577, + "step": 14610 + }, + { + "epoch": 0.45126242360639546, + "grad_norm": 0.37797215580940247, + "learning_rate": 9.00931735798141e-05, + "loss": 0.0526, + "step": 14620 + }, + { + "epoch": 0.451571084634854, + "grad_norm": 0.345033198595047, + "learning_rate": 9.00767014862654e-05, + "loss": 0.0509, + "step": 14630 + }, + { + "epoch": 0.45187974566331257, + "grad_norm": 0.33787018060684204, + "learning_rate": 9.006021721847295e-05, + "loss": 0.0555, + "step": 14640 + }, + { + "epoch": 0.45218840669177107, + "grad_norm": 0.3603231906890869, + "learning_rate": 9.004372078144423e-05, + "loss": 0.0508, + "step": 14650 + }, + { + "epoch": 0.45249706772022963, + "grad_norm": 0.34805238246917725, + "learning_rate": 9.002721218019043e-05, + "loss": 0.0553, + "step": 14660 + }, + { + "epoch": 0.4528057287486882, + "grad_norm": 0.5302698016166687, + "learning_rate": 9.001069141972642e-05, + "loss": 0.0595, + "step": 14670 + }, + { + "epoch": 0.45311438977714674, + "grad_norm": 0.3844645619392395, + "learning_rate": 8.99941585050708e-05, + "loss": 0.0525, + "step": 14680 + }, + { + "epoch": 0.4534230508056053, + "grad_norm": 0.26213186979293823, + "learning_rate": 8.997761344124578e-05, + "loss": 0.0491, + "step": 14690 + }, + { + "epoch": 0.45373171183406386, + "grad_norm": 0.38078293204307556, + "learning_rate": 8.996105623327737e-05, + "loss": 0.0582, + "step": 14700 + }, + { + "epoch": 0.45404037286252236, + "grad_norm": 0.46274319291114807, + "learning_rate": 8.994448688619517e-05, + "loss": 0.0502, + "step": 14710 + }, + { + "epoch": 0.4543490338909809, + "grad_norm": 0.9850357174873352, + "learning_rate": 8.992790540503253e-05, + "loss": 0.0541, + "step": 14720 + }, + { + "epoch": 0.45465769491943947, + "grad_norm": 0.40163955092430115, + "learning_rate": 8.991131179482648e-05, + "loss": 0.0573, + "step": 14730 + }, + { + "epoch": 0.454966355947898, + "grad_norm": 0.3970179557800293, + "learning_rate": 8.989470606061768e-05, + "loss": 0.0505, + "step": 14740 + }, + { + "epoch": 0.4552750169763566, + "grad_norm": 0.3043745160102844, + "learning_rate": 8.987808820745056e-05, + "loss": 0.0497, + "step": 14750 + }, + { + "epoch": 0.4555836780048151, + "grad_norm": 0.3856664001941681, + "learning_rate": 8.986145824037315e-05, + "loss": 0.056, + "step": 14760 + }, + { + "epoch": 0.45589233903327364, + "grad_norm": 0.2951919138431549, + "learning_rate": 8.984481616443721e-05, + "loss": 0.0506, + "step": 14770 + }, + { + "epoch": 0.4562010000617322, + "grad_norm": 0.3229632079601288, + "learning_rate": 8.982816198469815e-05, + "loss": 0.0528, + "step": 14780 + }, + { + "epoch": 0.45650966109019075, + "grad_norm": 0.30132901668548584, + "learning_rate": 8.98114957062151e-05, + "loss": 0.0503, + "step": 14790 + }, + { + "epoch": 0.4568183221186493, + "grad_norm": 0.359952837228775, + "learning_rate": 8.97948173340508e-05, + "loss": 0.0552, + "step": 14800 + }, + { + "epoch": 0.45712698314710787, + "grad_norm": 0.2149348258972168, + "learning_rate": 8.977812687327172e-05, + "loss": 0.0511, + "step": 14810 + }, + { + "epoch": 0.45743564417556637, + "grad_norm": 0.3752705454826355, + "learning_rate": 8.976142432894798e-05, + "loss": 0.0516, + "step": 14820 + }, + { + "epoch": 0.4577443052040249, + "grad_norm": 0.29930785298347473, + "learning_rate": 8.974470970615336e-05, + "loss": 0.0558, + "step": 14830 + }, + { + "epoch": 0.4580529662324835, + "grad_norm": 0.3683079481124878, + "learning_rate": 8.972798300996534e-05, + "loss": 0.0554, + "step": 14840 + }, + { + "epoch": 0.45836162726094204, + "grad_norm": 0.35566404461860657, + "learning_rate": 8.971124424546504e-05, + "loss": 0.0482, + "step": 14850 + }, + { + "epoch": 0.4586702882894006, + "grad_norm": 0.28047314286231995, + "learning_rate": 8.969449341773724e-05, + "loss": 0.0533, + "step": 14860 + }, + { + "epoch": 0.45897894931785915, + "grad_norm": 0.300666481256485, + "learning_rate": 8.967773053187042e-05, + "loss": 0.0555, + "step": 14870 + }, + { + "epoch": 0.45928761034631765, + "grad_norm": 0.4134526252746582, + "learning_rate": 8.966095559295668e-05, + "loss": 0.0534, + "step": 14880 + }, + { + "epoch": 0.4595962713747762, + "grad_norm": 0.26722100377082825, + "learning_rate": 8.964416860609184e-05, + "loss": 0.0492, + "step": 14890 + }, + { + "epoch": 0.45990493240323477, + "grad_norm": 0.5549184679985046, + "learning_rate": 8.962736957637532e-05, + "loss": 0.0566, + "step": 14900 + }, + { + "epoch": 0.4602135934316933, + "grad_norm": 1.5606201887130737, + "learning_rate": 8.96105585089102e-05, + "loss": 0.0563, + "step": 14910 + }, + { + "epoch": 0.4605222544601519, + "grad_norm": 0.3569463789463043, + "learning_rate": 8.959373540880329e-05, + "loss": 0.0533, + "step": 14920 + }, + { + "epoch": 0.4608309154886104, + "grad_norm": 0.375911682844162, + "learning_rate": 8.957690028116495e-05, + "loss": 0.0599, + "step": 14930 + }, + { + "epoch": 0.46113957651706894, + "grad_norm": 0.31800976395606995, + "learning_rate": 8.956005313110928e-05, + "loss": 0.0562, + "step": 14940 + }, + { + "epoch": 0.4614482375455275, + "grad_norm": 0.2550790011882782, + "learning_rate": 8.9543193963754e-05, + "loss": 0.0539, + "step": 14950 + }, + { + "epoch": 0.46175689857398605, + "grad_norm": 0.3639301061630249, + "learning_rate": 8.952632278422048e-05, + "loss": 0.0509, + "step": 14960 + }, + { + "epoch": 0.4620655596024446, + "grad_norm": 0.40793347358703613, + "learning_rate": 8.95094395976337e-05, + "loss": 0.0547, + "step": 14970 + }, + { + "epoch": 0.46237422063090317, + "grad_norm": 0.4125353991985321, + "learning_rate": 8.949254440912239e-05, + "loss": 0.0562, + "step": 14980 + }, + { + "epoch": 0.46268288165936167, + "grad_norm": 0.3278186321258545, + "learning_rate": 8.94756372238188e-05, + "loss": 0.0538, + "step": 14990 + }, + { + "epoch": 0.4629915426878202, + "grad_norm": 0.3780723512172699, + "learning_rate": 8.945871804685892e-05, + "loss": 0.0546, + "step": 15000 + }, + { + "epoch": 0.4633002037162788, + "grad_norm": 0.353526771068573, + "learning_rate": 8.944178688338236e-05, + "loss": 0.0575, + "step": 15010 + }, + { + "epoch": 0.46360886474473734, + "grad_norm": 0.2838698625564575, + "learning_rate": 8.942484373853233e-05, + "loss": 0.0581, + "step": 15020 + }, + { + "epoch": 0.4639175257731959, + "grad_norm": 0.3517088294029236, + "learning_rate": 8.940788861745572e-05, + "loss": 0.0515, + "step": 15030 + }, + { + "epoch": 0.46422618680165445, + "grad_norm": 0.35966289043426514, + "learning_rate": 8.939092152530308e-05, + "loss": 0.0537, + "step": 15040 + }, + { + "epoch": 0.46453484783011295, + "grad_norm": 0.27467429637908936, + "learning_rate": 8.937394246722853e-05, + "loss": 0.0519, + "step": 15050 + }, + { + "epoch": 0.4648435088585715, + "grad_norm": 0.3425252437591553, + "learning_rate": 8.935695144838984e-05, + "loss": 0.055, + "step": 15060 + }, + { + "epoch": 0.46515216988703006, + "grad_norm": 0.37264737486839294, + "learning_rate": 8.933994847394849e-05, + "loss": 0.0557, + "step": 15070 + }, + { + "epoch": 0.4654608309154886, + "grad_norm": 0.33186522126197815, + "learning_rate": 8.932293354906949e-05, + "loss": 0.0598, + "step": 15080 + }, + { + "epoch": 0.4657694919439472, + "grad_norm": 0.3879637122154236, + "learning_rate": 8.930590667892153e-05, + "loss": 0.0587, + "step": 15090 + }, + { + "epoch": 0.4660781529724057, + "grad_norm": 0.3386707305908203, + "learning_rate": 8.928886786867696e-05, + "loss": 0.054, + "step": 15100 + }, + { + "epoch": 0.46638681400086424, + "grad_norm": 0.42439451813697815, + "learning_rate": 8.927181712351168e-05, + "loss": 0.0572, + "step": 15110 + }, + { + "epoch": 0.4666954750293228, + "grad_norm": 0.34900808334350586, + "learning_rate": 8.925475444860527e-05, + "loss": 0.0517, + "step": 15120 + }, + { + "epoch": 0.46700413605778135, + "grad_norm": 0.5109984874725342, + "learning_rate": 8.923767984914092e-05, + "loss": 0.0488, + "step": 15130 + }, + { + "epoch": 0.4673127970862399, + "grad_norm": 0.30740052461624146, + "learning_rate": 8.922059333030545e-05, + "loss": 0.0547, + "step": 15140 + }, + { + "epoch": 0.46762145811469846, + "grad_norm": 0.2684696912765503, + "learning_rate": 8.920349489728928e-05, + "loss": 0.0516, + "step": 15150 + }, + { + "epoch": 0.46793011914315696, + "grad_norm": 0.30346783995628357, + "learning_rate": 8.918638455528646e-05, + "loss": 0.0518, + "step": 15160 + }, + { + "epoch": 0.4682387801716155, + "grad_norm": 0.3005162477493286, + "learning_rate": 8.916926230949468e-05, + "loss": 0.0536, + "step": 15170 + }, + { + "epoch": 0.4685474412000741, + "grad_norm": 0.3024720251560211, + "learning_rate": 8.915212816511522e-05, + "loss": 0.0509, + "step": 15180 + }, + { + "epoch": 0.46885610222853263, + "grad_norm": 0.46665361523628235, + "learning_rate": 8.913498212735296e-05, + "loss": 0.0585, + "step": 15190 + }, + { + "epoch": 0.4691647632569912, + "grad_norm": 0.5301529765129089, + "learning_rate": 8.911782420141643e-05, + "loss": 0.0521, + "step": 15200 + }, + { + "epoch": 0.4694734242854497, + "grad_norm": 0.3267645239830017, + "learning_rate": 8.910065439251775e-05, + "loss": 0.0555, + "step": 15210 + }, + { + "epoch": 0.46978208531390825, + "grad_norm": 0.4809127151966095, + "learning_rate": 8.908347270587268e-05, + "loss": 0.0554, + "step": 15220 + }, + { + "epoch": 0.4700907463423668, + "grad_norm": 0.2926150858402252, + "learning_rate": 8.906627914670054e-05, + "loss": 0.0539, + "step": 15230 + }, + { + "epoch": 0.47039940737082536, + "grad_norm": 0.41854041814804077, + "learning_rate": 8.904907372022427e-05, + "loss": 0.0537, + "step": 15240 + }, + { + "epoch": 0.4707080683992839, + "grad_norm": 0.3459557592868805, + "learning_rate": 8.903185643167042e-05, + "loss": 0.055, + "step": 15250 + }, + { + "epoch": 0.4710167294277425, + "grad_norm": 0.24701935052871704, + "learning_rate": 8.901462728626919e-05, + "loss": 0.0515, + "step": 15260 + }, + { + "epoch": 0.471325390456201, + "grad_norm": 0.2862972021102905, + "learning_rate": 8.899738628925429e-05, + "loss": 0.0584, + "step": 15270 + }, + { + "epoch": 0.47163405148465953, + "grad_norm": 0.2616805136203766, + "learning_rate": 8.898013344586312e-05, + "loss": 0.0545, + "step": 15280 + }, + { + "epoch": 0.4719427125131181, + "grad_norm": 0.24263639748096466, + "learning_rate": 8.896286876133661e-05, + "loss": 0.0547, + "step": 15290 + }, + { + "epoch": 0.47225137354157665, + "grad_norm": 0.3991931676864624, + "learning_rate": 8.894559224091933e-05, + "loss": 0.0552, + "step": 15300 + }, + { + "epoch": 0.4725600345700352, + "grad_norm": 0.7607799768447876, + "learning_rate": 8.892830388985942e-05, + "loss": 0.049, + "step": 15310 + }, + { + "epoch": 0.47286869559849376, + "grad_norm": 0.2940968871116638, + "learning_rate": 8.891100371340864e-05, + "loss": 0.0484, + "step": 15320 + }, + { + "epoch": 0.47317735662695226, + "grad_norm": 0.3282192647457123, + "learning_rate": 8.889369171682231e-05, + "loss": 0.0524, + "step": 15330 + }, + { + "epoch": 0.4734860176554108, + "grad_norm": 0.3039143979549408, + "learning_rate": 8.887636790535936e-05, + "loss": 0.0532, + "step": 15340 + }, + { + "epoch": 0.4737946786838694, + "grad_norm": 0.3284814953804016, + "learning_rate": 8.885903228428231e-05, + "loss": 0.0493, + "step": 15350 + }, + { + "epoch": 0.47410333971232793, + "grad_norm": 0.3960978090763092, + "learning_rate": 8.884168485885727e-05, + "loss": 0.0527, + "step": 15360 + }, + { + "epoch": 0.4744120007407865, + "grad_norm": 0.28426775336265564, + "learning_rate": 8.882432563435393e-05, + "loss": 0.0578, + "step": 15370 + }, + { + "epoch": 0.474720661769245, + "grad_norm": 0.34814149141311646, + "learning_rate": 8.880695461604556e-05, + "loss": 0.0539, + "step": 15380 + }, + { + "epoch": 0.47502932279770355, + "grad_norm": 0.2988288104534149, + "learning_rate": 8.878957180920901e-05, + "loss": 0.0548, + "step": 15390 + }, + { + "epoch": 0.4753379838261621, + "grad_norm": 0.4530220329761505, + "learning_rate": 8.877217721912473e-05, + "loss": 0.0538, + "step": 15400 + }, + { + "epoch": 0.47564664485462066, + "grad_norm": 0.3038022816181183, + "learning_rate": 8.875477085107673e-05, + "loss": 0.0477, + "step": 15410 + }, + { + "epoch": 0.4759553058830792, + "grad_norm": 0.3802969753742218, + "learning_rate": 8.87373527103526e-05, + "loss": 0.0574, + "step": 15420 + }, + { + "epoch": 0.4762639669115378, + "grad_norm": 0.3184208869934082, + "learning_rate": 8.871992280224353e-05, + "loss": 0.0519, + "step": 15430 + }, + { + "epoch": 0.4765726279399963, + "grad_norm": 0.3091546297073364, + "learning_rate": 8.870248113204422e-05, + "loss": 0.0547, + "step": 15440 + }, + { + "epoch": 0.47688128896845483, + "grad_norm": 0.29328182339668274, + "learning_rate": 8.868502770505306e-05, + "loss": 0.0486, + "step": 15450 + }, + { + "epoch": 0.4771899499969134, + "grad_norm": 0.44737759232521057, + "learning_rate": 8.86675625265719e-05, + "loss": 0.0531, + "step": 15460 + }, + { + "epoch": 0.47749861102537194, + "grad_norm": 0.4545552134513855, + "learning_rate": 8.865008560190618e-05, + "loss": 0.0496, + "step": 15470 + }, + { + "epoch": 0.4778072720538305, + "grad_norm": 0.2548025846481323, + "learning_rate": 8.863259693636496e-05, + "loss": 0.053, + "step": 15480 + }, + { + "epoch": 0.47811593308228906, + "grad_norm": 0.32714107632637024, + "learning_rate": 8.861509653526083e-05, + "loss": 0.0557, + "step": 15490 + }, + { + "epoch": 0.47842459411074756, + "grad_norm": 0.6294752359390259, + "learning_rate": 8.859758440390993e-05, + "loss": 0.0542, + "step": 15500 + }, + { + "epoch": 0.4787332551392061, + "grad_norm": 0.315367192029953, + "learning_rate": 8.858006054763202e-05, + "loss": 0.0535, + "step": 15510 + }, + { + "epoch": 0.47904191616766467, + "grad_norm": 0.36701899766921997, + "learning_rate": 8.856252497175035e-05, + "loss": 0.0537, + "step": 15520 + }, + { + "epoch": 0.47935057719612323, + "grad_norm": 0.3528570532798767, + "learning_rate": 8.854497768159178e-05, + "loss": 0.0483, + "step": 15530 + }, + { + "epoch": 0.4796592382245818, + "grad_norm": 0.25059792399406433, + "learning_rate": 8.852741868248671e-05, + "loss": 0.0553, + "step": 15540 + }, + { + "epoch": 0.4799678992530403, + "grad_norm": 0.40667062997817993, + "learning_rate": 8.85098479797691e-05, + "loss": 0.0556, + "step": 15550 + }, + { + "epoch": 0.48027656028149884, + "grad_norm": 0.3144872188568115, + "learning_rate": 8.849226557877646e-05, + "loss": 0.0553, + "step": 15560 + }, + { + "epoch": 0.4805852213099574, + "grad_norm": 0.34852859377861023, + "learning_rate": 8.84746714848499e-05, + "loss": 0.0567, + "step": 15570 + }, + { + "epoch": 0.48089388233841596, + "grad_norm": 0.3891640901565552, + "learning_rate": 8.845706570333397e-05, + "loss": 0.0558, + "step": 15580 + }, + { + "epoch": 0.4812025433668745, + "grad_norm": 0.356802374124527, + "learning_rate": 8.84394482395769e-05, + "loss": 0.0526, + "step": 15590 + }, + { + "epoch": 0.48151120439533307, + "grad_norm": 0.3529358506202698, + "learning_rate": 8.842181909893038e-05, + "loss": 0.0523, + "step": 15600 + }, + { + "epoch": 0.48181986542379157, + "grad_norm": 0.3542776107788086, + "learning_rate": 8.840417828674969e-05, + "loss": 0.0546, + "step": 15610 + }, + { + "epoch": 0.48212852645225013, + "grad_norm": 0.4084323048591614, + "learning_rate": 8.838652580839364e-05, + "loss": 0.0503, + "step": 15620 + }, + { + "epoch": 0.4824371874807087, + "grad_norm": 0.4297856390476227, + "learning_rate": 8.836886166922458e-05, + "loss": 0.052, + "step": 15630 + }, + { + "epoch": 0.48274584850916724, + "grad_norm": 0.3210156559944153, + "learning_rate": 8.835118587460844e-05, + "loss": 0.0504, + "step": 15640 + }, + { + "epoch": 0.4830545095376258, + "grad_norm": 0.5510222911834717, + "learning_rate": 8.83334984299146e-05, + "loss": 0.0513, + "step": 15650 + }, + { + "epoch": 0.4833631705660843, + "grad_norm": 0.2920938730239868, + "learning_rate": 8.83157993405161e-05, + "loss": 0.0542, + "step": 15660 + }, + { + "epoch": 0.48367183159454286, + "grad_norm": 0.2978890538215637, + "learning_rate": 8.829808861178943e-05, + "loss": 0.05, + "step": 15670 + }, + { + "epoch": 0.4839804926230014, + "grad_norm": 0.3313884139060974, + "learning_rate": 8.828036624911464e-05, + "loss": 0.0522, + "step": 15680 + }, + { + "epoch": 0.48428915365145997, + "grad_norm": 0.2573363482952118, + "learning_rate": 8.826263225787532e-05, + "loss": 0.0501, + "step": 15690 + }, + { + "epoch": 0.4845978146799185, + "grad_norm": 0.3287264108657837, + "learning_rate": 8.824488664345858e-05, + "loss": 0.0523, + "step": 15700 + }, + { + "epoch": 0.4849064757083771, + "grad_norm": 0.4203445017337799, + "learning_rate": 8.822712941125508e-05, + "loss": 0.0551, + "step": 15710 + }, + { + "epoch": 0.4852151367368356, + "grad_norm": 0.3276310861110687, + "learning_rate": 8.820936056665898e-05, + "loss": 0.0525, + "step": 15720 + }, + { + "epoch": 0.48552379776529414, + "grad_norm": 0.2728821039199829, + "learning_rate": 8.819158011506801e-05, + "loss": 0.0492, + "step": 15730 + }, + { + "epoch": 0.4858324587937527, + "grad_norm": 0.31726348400115967, + "learning_rate": 8.81737880618834e-05, + "loss": 0.049, + "step": 15740 + }, + { + "epoch": 0.48614111982221125, + "grad_norm": 0.3508398234844208, + "learning_rate": 8.815598441250987e-05, + "loss": 0.0547, + "step": 15750 + }, + { + "epoch": 0.4864497808506698, + "grad_norm": 0.44309234619140625, + "learning_rate": 8.813816917235576e-05, + "loss": 0.0522, + "step": 15760 + }, + { + "epoch": 0.48675844187912837, + "grad_norm": 0.6359990239143372, + "learning_rate": 8.812034234683282e-05, + "loss": 0.0555, + "step": 15770 + }, + { + "epoch": 0.48706710290758687, + "grad_norm": 0.38196927309036255, + "learning_rate": 8.810250394135637e-05, + "loss": 0.0548, + "step": 15780 + }, + { + "epoch": 0.4873757639360454, + "grad_norm": 0.6302274465560913, + "learning_rate": 8.808465396134529e-05, + "loss": 0.0603, + "step": 15790 + }, + { + "epoch": 0.487684424964504, + "grad_norm": 0.421142041683197, + "learning_rate": 8.806679241222189e-05, + "loss": 0.0541, + "step": 15800 + }, + { + "epoch": 0.48799308599296254, + "grad_norm": 0.29614418745040894, + "learning_rate": 8.804891929941203e-05, + "loss": 0.0535, + "step": 15810 + }, + { + "epoch": 0.4883017470214211, + "grad_norm": 0.34537336230278015, + "learning_rate": 8.803103462834514e-05, + "loss": 0.0522, + "step": 15820 + }, + { + "epoch": 0.4886104080498796, + "grad_norm": 0.3951790928840637, + "learning_rate": 8.801313840445408e-05, + "loss": 0.0515, + "step": 15830 + }, + { + "epoch": 0.48891906907833815, + "grad_norm": 0.38858479261398315, + "learning_rate": 8.799523063317524e-05, + "loss": 0.0533, + "step": 15840 + }, + { + "epoch": 0.4892277301067967, + "grad_norm": 0.30676499009132385, + "learning_rate": 8.797731131994854e-05, + "loss": 0.0536, + "step": 15850 + }, + { + "epoch": 0.48953639113525527, + "grad_norm": 0.30029839277267456, + "learning_rate": 8.795938047021739e-05, + "loss": 0.0547, + "step": 15860 + }, + { + "epoch": 0.4898450521637138, + "grad_norm": 0.44375917315483093, + "learning_rate": 8.794143808942872e-05, + "loss": 0.0554, + "step": 15870 + }, + { + "epoch": 0.4901537131921724, + "grad_norm": 0.27715975046157837, + "learning_rate": 8.792348418303296e-05, + "loss": 0.054, + "step": 15880 + }, + { + "epoch": 0.4904623742206309, + "grad_norm": 0.3246401250362396, + "learning_rate": 8.790551875648398e-05, + "loss": 0.0605, + "step": 15890 + }, + { + "epoch": 0.49077103524908944, + "grad_norm": 0.28742268681526184, + "learning_rate": 8.788754181523926e-05, + "loss": 0.0504, + "step": 15900 + }, + { + "epoch": 0.491079696277548, + "grad_norm": 0.2928858697414398, + "learning_rate": 8.78695533647597e-05, + "loss": 0.0559, + "step": 15910 + }, + { + "epoch": 0.49138835730600655, + "grad_norm": 0.3906520903110504, + "learning_rate": 8.785155341050972e-05, + "loss": 0.0512, + "step": 15920 + }, + { + "epoch": 0.4916970183344651, + "grad_norm": 0.5193008184432983, + "learning_rate": 8.783354195795721e-05, + "loss": 0.0524, + "step": 15930 + }, + { + "epoch": 0.49200567936292366, + "grad_norm": 0.3068596422672272, + "learning_rate": 8.78155190125736e-05, + "loss": 0.0498, + "step": 15940 + }, + { + "epoch": 0.49231434039138217, + "grad_norm": 0.2979864478111267, + "learning_rate": 8.779748457983378e-05, + "loss": 0.0561, + "step": 15950 + }, + { + "epoch": 0.4926230014198407, + "grad_norm": 0.31987300515174866, + "learning_rate": 8.777943866521612e-05, + "loss": 0.0506, + "step": 15960 + }, + { + "epoch": 0.4929316624482993, + "grad_norm": 0.3047407269477844, + "learning_rate": 8.77613812742025e-05, + "loss": 0.0551, + "step": 15970 + }, + { + "epoch": 0.49324032347675784, + "grad_norm": 0.29245010018348694, + "learning_rate": 8.774331241227829e-05, + "loss": 0.057, + "step": 15980 + }, + { + "epoch": 0.4935489845052164, + "grad_norm": 0.39124566316604614, + "learning_rate": 8.772523208493232e-05, + "loss": 0.0506, + "step": 15990 + }, + { + "epoch": 0.4938576455336749, + "grad_norm": 0.34273695945739746, + "learning_rate": 8.770714029765692e-05, + "loss": 0.0508, + "step": 16000 + }, + { + "epoch": 0.49416630656213345, + "grad_norm": 0.33422350883483887, + "learning_rate": 8.768903705594789e-05, + "loss": 0.0497, + "step": 16010 + }, + { + "epoch": 0.494474967590592, + "grad_norm": 0.29569458961486816, + "learning_rate": 8.767092236530453e-05, + "loss": 0.0546, + "step": 16020 + }, + { + "epoch": 0.49478362861905056, + "grad_norm": 0.935872495174408, + "learning_rate": 8.76527962312296e-05, + "loss": 0.0484, + "step": 16030 + }, + { + "epoch": 0.4950922896475091, + "grad_norm": 0.32451480627059937, + "learning_rate": 8.763465865922934e-05, + "loss": 0.0506, + "step": 16040 + }, + { + "epoch": 0.4954009506759677, + "grad_norm": 0.3407304584980011, + "learning_rate": 8.761650965481347e-05, + "loss": 0.0504, + "step": 16050 + }, + { + "epoch": 0.4957096117044262, + "grad_norm": 0.6772223711013794, + "learning_rate": 8.759834922349516e-05, + "loss": 0.0575, + "step": 16060 + }, + { + "epoch": 0.49601827273288474, + "grad_norm": 1.1094279289245605, + "learning_rate": 8.758017737079108e-05, + "loss": 0.0506, + "step": 16070 + }, + { + "epoch": 0.4963269337613433, + "grad_norm": 0.5693839192390442, + "learning_rate": 8.756199410222137e-05, + "loss": 0.0535, + "step": 16080 + }, + { + "epoch": 0.49663559478980185, + "grad_norm": 0.3497790992259979, + "learning_rate": 8.754379942330963e-05, + "loss": 0.0578, + "step": 16090 + }, + { + "epoch": 0.4969442558182604, + "grad_norm": 0.4990483522415161, + "learning_rate": 8.75255933395829e-05, + "loss": 0.0445, + "step": 16100 + }, + { + "epoch": 0.4972529168467189, + "grad_norm": 0.377096951007843, + "learning_rate": 8.750737585657171e-05, + "loss": 0.053, + "step": 16110 + }, + { + "epoch": 0.49756157787517746, + "grad_norm": 0.34984126687049866, + "learning_rate": 8.748914697981008e-05, + "loss": 0.053, + "step": 16120 + }, + { + "epoch": 0.497870238903636, + "grad_norm": 0.31372156739234924, + "learning_rate": 8.747090671483542e-05, + "loss": 0.0514, + "step": 16130 + }, + { + "epoch": 0.4981788999320946, + "grad_norm": 0.27952343225479126, + "learning_rate": 8.745265506718869e-05, + "loss": 0.0541, + "step": 16140 + }, + { + "epoch": 0.49848756096055313, + "grad_norm": 0.5177311897277832, + "learning_rate": 8.74343920424142e-05, + "loss": 0.0552, + "step": 16150 + }, + { + "epoch": 0.4987962219890117, + "grad_norm": 0.24799217283725739, + "learning_rate": 8.741611764605982e-05, + "loss": 0.0523, + "step": 16160 + }, + { + "epoch": 0.4991048830174702, + "grad_norm": 0.37201032042503357, + "learning_rate": 8.739783188367682e-05, + "loss": 0.0557, + "step": 16170 + }, + { + "epoch": 0.49941354404592875, + "grad_norm": 0.4090786874294281, + "learning_rate": 8.737953476081991e-05, + "loss": 0.0526, + "step": 16180 + }, + { + "epoch": 0.4997222050743873, + "grad_norm": 0.3231889605522156, + "learning_rate": 8.73612262830473e-05, + "loss": 0.0551, + "step": 16190 + }, + { + "epoch": 0.5000308661028459, + "grad_norm": 0.39312848448753357, + "learning_rate": 8.734290645592061e-05, + "loss": 0.0568, + "step": 16200 + }, + { + "epoch": 0.5003395271313044, + "grad_norm": 0.32627207040786743, + "learning_rate": 8.732457528500493e-05, + "loss": 0.0553, + "step": 16210 + }, + { + "epoch": 0.500648188159763, + "grad_norm": 0.27325737476348877, + "learning_rate": 8.730623277586875e-05, + "loss": 0.0575, + "step": 16220 + }, + { + "epoch": 0.5009568491882215, + "grad_norm": 0.290724515914917, + "learning_rate": 8.72878789340841e-05, + "loss": 0.048, + "step": 16230 + }, + { + "epoch": 0.5012655102166801, + "grad_norm": 0.36230233311653137, + "learning_rate": 8.726951376522635e-05, + "loss": 0.0508, + "step": 16240 + }, + { + "epoch": 0.5015741712451386, + "grad_norm": 0.324452668428421, + "learning_rate": 8.725113727487435e-05, + "loss": 0.0553, + "step": 16250 + }, + { + "epoch": 0.5018828322735971, + "grad_norm": 0.5025396943092346, + "learning_rate": 8.723274946861042e-05, + "loss": 0.0536, + "step": 16260 + }, + { + "epoch": 0.5021914933020557, + "grad_norm": 0.3688121736049652, + "learning_rate": 8.721435035202026e-05, + "loss": 0.057, + "step": 16270 + }, + { + "epoch": 0.5025001543305142, + "grad_norm": 0.5798941254615784, + "learning_rate": 8.719593993069306e-05, + "loss": 0.0561, + "step": 16280 + }, + { + "epoch": 0.5028088153589728, + "grad_norm": 0.3501107692718506, + "learning_rate": 8.717751821022139e-05, + "loss": 0.0613, + "step": 16290 + }, + { + "epoch": 0.5031174763874313, + "grad_norm": 0.32255083322525024, + "learning_rate": 8.715908519620134e-05, + "loss": 0.0566, + "step": 16300 + }, + { + "epoch": 0.5034261374158898, + "grad_norm": 0.40577858686447144, + "learning_rate": 8.71406408942323e-05, + "loss": 0.0531, + "step": 16310 + }, + { + "epoch": 0.5037347984443484, + "grad_norm": 0.27784404158592224, + "learning_rate": 8.712218530991723e-05, + "loss": 0.0492, + "step": 16320 + }, + { + "epoch": 0.5040434594728069, + "grad_norm": 0.49037790298461914, + "learning_rate": 8.710371844886241e-05, + "loss": 0.0487, + "step": 16330 + }, + { + "epoch": 0.5043521205012655, + "grad_norm": 0.42779627442359924, + "learning_rate": 8.708524031667758e-05, + "loss": 0.0548, + "step": 16340 + }, + { + "epoch": 0.504660781529724, + "grad_norm": 0.3688235580921173, + "learning_rate": 8.706675091897592e-05, + "loss": 0.0513, + "step": 16350 + }, + { + "epoch": 0.5049694425581827, + "grad_norm": 0.404933899641037, + "learning_rate": 8.704825026137404e-05, + "loss": 0.0531, + "step": 16360 + }, + { + "epoch": 0.5052781035866412, + "grad_norm": 0.39559412002563477, + "learning_rate": 8.702973834949192e-05, + "loss": 0.0511, + "step": 16370 + }, + { + "epoch": 0.5055867646150997, + "grad_norm": 0.283434122800827, + "learning_rate": 8.701121518895301e-05, + "loss": 0.0513, + "step": 16380 + }, + { + "epoch": 0.5058954256435583, + "grad_norm": 0.53680819272995, + "learning_rate": 8.699268078538414e-05, + "loss": 0.0579, + "step": 16390 + }, + { + "epoch": 0.5062040866720168, + "grad_norm": 0.23742596805095673, + "learning_rate": 8.69741351444156e-05, + "loss": 0.0515, + "step": 16400 + }, + { + "epoch": 0.5065127477004754, + "grad_norm": 0.2979493737220764, + "learning_rate": 8.695557827168101e-05, + "loss": 0.0539, + "step": 16410 + }, + { + "epoch": 0.5068214087289339, + "grad_norm": 0.32941171526908875, + "learning_rate": 8.693701017281753e-05, + "loss": 0.0537, + "step": 16420 + }, + { + "epoch": 0.5071300697573924, + "grad_norm": 1.1365513801574707, + "learning_rate": 8.691843085346563e-05, + "loss": 0.0527, + "step": 16430 + }, + { + "epoch": 0.507438730785851, + "grad_norm": 0.4139597713947296, + "learning_rate": 8.689984031926919e-05, + "loss": 0.0533, + "step": 16440 + }, + { + "epoch": 0.5077473918143095, + "grad_norm": 0.3216039538383484, + "learning_rate": 8.688123857587555e-05, + "loss": 0.0536, + "step": 16450 + }, + { + "epoch": 0.5080560528427681, + "grad_norm": 0.2875566780567169, + "learning_rate": 8.686262562893544e-05, + "loss": 0.0475, + "step": 16460 + }, + { + "epoch": 0.5083647138712266, + "grad_norm": 0.3247299790382385, + "learning_rate": 8.684400148410294e-05, + "loss": 0.0519, + "step": 16470 + }, + { + "epoch": 0.5086733748996851, + "grad_norm": 0.43387526273727417, + "learning_rate": 8.682536614703562e-05, + "loss": 0.0544, + "step": 16480 + }, + { + "epoch": 0.5089820359281437, + "grad_norm": 0.27847498655319214, + "learning_rate": 8.680671962339437e-05, + "loss": 0.054, + "step": 16490 + }, + { + "epoch": 0.5092906969566022, + "grad_norm": 0.3808903694152832, + "learning_rate": 8.678806191884352e-05, + "loss": 0.0491, + "step": 16500 + }, + { + "epoch": 0.5095993579850608, + "grad_norm": 0.33295938372612, + "learning_rate": 8.67693930390508e-05, + "loss": 0.0504, + "step": 16510 + }, + { + "epoch": 0.5099080190135193, + "grad_norm": 0.42129412293434143, + "learning_rate": 8.67507129896873e-05, + "loss": 0.0521, + "step": 16520 + }, + { + "epoch": 0.510216680041978, + "grad_norm": 0.2896689176559448, + "learning_rate": 8.673202177642757e-05, + "loss": 0.0511, + "step": 16530 + }, + { + "epoch": 0.5105253410704365, + "grad_norm": 0.3296789526939392, + "learning_rate": 8.671331940494945e-05, + "loss": 0.0488, + "step": 16540 + }, + { + "epoch": 0.510834002098895, + "grad_norm": 0.344502329826355, + "learning_rate": 8.669460588093427e-05, + "loss": 0.0484, + "step": 16550 + }, + { + "epoch": 0.5111426631273536, + "grad_norm": 0.335074782371521, + "learning_rate": 8.667588121006667e-05, + "loss": 0.052, + "step": 16560 + }, + { + "epoch": 0.5114513241558121, + "grad_norm": 0.32395312190055847, + "learning_rate": 8.665714539803475e-05, + "loss": 0.0513, + "step": 16570 + }, + { + "epoch": 0.5117599851842707, + "grad_norm": 0.26896932721138, + "learning_rate": 8.663839845052993e-05, + "loss": 0.052, + "step": 16580 + }, + { + "epoch": 0.5120686462127292, + "grad_norm": 0.30801600217819214, + "learning_rate": 8.661964037324703e-05, + "loss": 0.053, + "step": 16590 + }, + { + "epoch": 0.5123773072411877, + "grad_norm": 0.3840690851211548, + "learning_rate": 8.660087117188427e-05, + "loss": 0.054, + "step": 16600 + }, + { + "epoch": 0.5126859682696463, + "grad_norm": 0.4175031781196594, + "learning_rate": 8.658209085214325e-05, + "loss": 0.0509, + "step": 16610 + }, + { + "epoch": 0.5129946292981048, + "grad_norm": 0.3571467697620392, + "learning_rate": 8.656329941972891e-05, + "loss": 0.0498, + "step": 16620 + }, + { + "epoch": 0.5133032903265634, + "grad_norm": 0.38345828652381897, + "learning_rate": 8.654449688034963e-05, + "loss": 0.0517, + "step": 16630 + }, + { + "epoch": 0.5136119513550219, + "grad_norm": 0.30289337038993835, + "learning_rate": 8.652568323971706e-05, + "loss": 0.0555, + "step": 16640 + }, + { + "epoch": 0.5139206123834804, + "grad_norm": 0.3041256368160248, + "learning_rate": 8.650685850354636e-05, + "loss": 0.0506, + "step": 16650 + }, + { + "epoch": 0.514229273411939, + "grad_norm": 0.2633620500564575, + "learning_rate": 8.648802267755593e-05, + "loss": 0.0506, + "step": 16660 + }, + { + "epoch": 0.5145379344403975, + "grad_norm": 0.327226847410202, + "learning_rate": 8.646917576746764e-05, + "loss": 0.0556, + "step": 16670 + }, + { + "epoch": 0.5148465954688561, + "grad_norm": 0.5412476062774658, + "learning_rate": 8.645031777900666e-05, + "loss": 0.0549, + "step": 16680 + }, + { + "epoch": 0.5151552564973146, + "grad_norm": 0.3562469780445099, + "learning_rate": 8.643144871790154e-05, + "loss": 0.0602, + "step": 16690 + }, + { + "epoch": 0.5154639175257731, + "grad_norm": 0.3046448230743408, + "learning_rate": 8.641256858988424e-05, + "loss": 0.0508, + "step": 16700 + }, + { + "epoch": 0.5157725785542318, + "grad_norm": 0.5161900520324707, + "learning_rate": 8.639367740069e-05, + "loss": 0.0503, + "step": 16710 + }, + { + "epoch": 0.5160812395826903, + "grad_norm": 0.36111775040626526, + "learning_rate": 8.63747751560575e-05, + "loss": 0.0482, + "step": 16720 + }, + { + "epoch": 0.5163899006111489, + "grad_norm": 0.37005138397216797, + "learning_rate": 8.635586186172871e-05, + "loss": 0.0512, + "step": 16730 + }, + { + "epoch": 0.5166985616396074, + "grad_norm": 0.6963960528373718, + "learning_rate": 8.633693752344902e-05, + "loss": 0.0511, + "step": 16740 + }, + { + "epoch": 0.517007222668066, + "grad_norm": 0.35088253021240234, + "learning_rate": 8.631800214696713e-05, + "loss": 0.0531, + "step": 16750 + }, + { + "epoch": 0.5173158836965245, + "grad_norm": 0.42901474237442017, + "learning_rate": 8.629905573803511e-05, + "loss": 0.053, + "step": 16760 + }, + { + "epoch": 0.517624544724983, + "grad_norm": 0.42774420976638794, + "learning_rate": 8.628009830240839e-05, + "loss": 0.0534, + "step": 16770 + }, + { + "epoch": 0.5179332057534416, + "grad_norm": 0.3832734227180481, + "learning_rate": 8.626112984584571e-05, + "loss": 0.0515, + "step": 16780 + }, + { + "epoch": 0.5182418667819001, + "grad_norm": 0.30487048625946045, + "learning_rate": 8.62421503741092e-05, + "loss": 0.0527, + "step": 16790 + }, + { + "epoch": 0.5185505278103587, + "grad_norm": 0.2766411304473877, + "learning_rate": 8.622315989296432e-05, + "loss": 0.0563, + "step": 16800 + }, + { + "epoch": 0.5188591888388172, + "grad_norm": 0.572365939617157, + "learning_rate": 8.62041584081799e-05, + "loss": 0.0539, + "step": 16810 + }, + { + "epoch": 0.5191678498672757, + "grad_norm": 0.3819721043109894, + "learning_rate": 8.618514592552807e-05, + "loss": 0.0483, + "step": 16820 + }, + { + "epoch": 0.5194765108957343, + "grad_norm": 0.3682754635810852, + "learning_rate": 8.616612245078431e-05, + "loss": 0.0532, + "step": 16830 + }, + { + "epoch": 0.5197851719241928, + "grad_norm": 0.2861032485961914, + "learning_rate": 8.614708798972746e-05, + "loss": 0.0523, + "step": 16840 + }, + { + "epoch": 0.5200938329526514, + "grad_norm": 0.3219168782234192, + "learning_rate": 8.61280425481397e-05, + "loss": 0.0521, + "step": 16850 + }, + { + "epoch": 0.5204024939811099, + "grad_norm": 0.3370620310306549, + "learning_rate": 8.61089861318065e-05, + "loss": 0.0509, + "step": 16860 + }, + { + "epoch": 0.5207111550095684, + "grad_norm": 0.32443398237228394, + "learning_rate": 8.608991874651673e-05, + "loss": 0.0529, + "step": 16870 + }, + { + "epoch": 0.521019816038027, + "grad_norm": 0.4291456937789917, + "learning_rate": 8.607084039806255e-05, + "loss": 0.0527, + "step": 16880 + }, + { + "epoch": 0.5213284770664856, + "grad_norm": 0.24520064890384674, + "learning_rate": 8.605175109223944e-05, + "loss": 0.0514, + "step": 16890 + }, + { + "epoch": 0.5216371380949442, + "grad_norm": 0.33488181233406067, + "learning_rate": 8.603265083484624e-05, + "loss": 0.0519, + "step": 16900 + }, + { + "epoch": 0.5219457991234027, + "grad_norm": 0.2882262170314789, + "learning_rate": 8.60135396316851e-05, + "loss": 0.0523, + "step": 16910 + }, + { + "epoch": 0.5222544601518613, + "grad_norm": 0.31198787689208984, + "learning_rate": 8.599441748856152e-05, + "loss": 0.0514, + "step": 16920 + }, + { + "epoch": 0.5225631211803198, + "grad_norm": 0.5634285807609558, + "learning_rate": 8.597528441128427e-05, + "loss": 0.0523, + "step": 16930 + }, + { + "epoch": 0.5228717822087783, + "grad_norm": 0.26500195264816284, + "learning_rate": 8.595614040566549e-05, + "loss": 0.0492, + "step": 16940 + }, + { + "epoch": 0.5231804432372369, + "grad_norm": 0.31138619780540466, + "learning_rate": 8.593698547752063e-05, + "loss": 0.0499, + "step": 16950 + }, + { + "epoch": 0.5234891042656954, + "grad_norm": 0.6272678971290588, + "learning_rate": 8.591781963266843e-05, + "loss": 0.0508, + "step": 16960 + }, + { + "epoch": 0.523797765294154, + "grad_norm": 0.3462122082710266, + "learning_rate": 8.5898642876931e-05, + "loss": 0.0472, + "step": 16970 + }, + { + "epoch": 0.5241064263226125, + "grad_norm": 0.3438381552696228, + "learning_rate": 8.587945521613369e-05, + "loss": 0.0511, + "step": 16980 + }, + { + "epoch": 0.524415087351071, + "grad_norm": 0.38226190209388733, + "learning_rate": 8.586025665610524e-05, + "loss": 0.0545, + "step": 16990 + }, + { + "epoch": 0.5247237483795296, + "grad_norm": 0.3254014849662781, + "learning_rate": 8.584104720267765e-05, + "loss": 0.0535, + "step": 17000 + }, + { + "epoch": 0.5250324094079881, + "grad_norm": 0.3573308289051056, + "learning_rate": 8.582182686168625e-05, + "loss": 0.0561, + "step": 17010 + }, + { + "epoch": 0.5253410704364467, + "grad_norm": 0.6978538632392883, + "learning_rate": 8.580259563896967e-05, + "loss": 0.0506, + "step": 17020 + }, + { + "epoch": 0.5256497314649052, + "grad_norm": 0.25046685338020325, + "learning_rate": 8.578335354036983e-05, + "loss": 0.0511, + "step": 17030 + }, + { + "epoch": 0.5259583924933637, + "grad_norm": 0.3178999125957489, + "learning_rate": 8.576410057173201e-05, + "loss": 0.0514, + "step": 17040 + }, + { + "epoch": 0.5262670535218223, + "grad_norm": 0.2989979684352875, + "learning_rate": 8.574483673890474e-05, + "loss": 0.051, + "step": 17050 + }, + { + "epoch": 0.5265757145502808, + "grad_norm": 0.4534737169742584, + "learning_rate": 8.572556204773983e-05, + "loss": 0.0536, + "step": 17060 + }, + { + "epoch": 0.5268843755787395, + "grad_norm": 0.3415033221244812, + "learning_rate": 8.570627650409246e-05, + "loss": 0.0457, + "step": 17070 + }, + { + "epoch": 0.527193036607198, + "grad_norm": 0.4052601754665375, + "learning_rate": 8.568698011382107e-05, + "loss": 0.0518, + "step": 17080 + }, + { + "epoch": 0.5275016976356566, + "grad_norm": 0.3424273431301117, + "learning_rate": 8.566767288278738e-05, + "loss": 0.0512, + "step": 17090 + }, + { + "epoch": 0.5278103586641151, + "grad_norm": 0.2831941246986389, + "learning_rate": 8.56483548168564e-05, + "loss": 0.0523, + "step": 17100 + }, + { + "epoch": 0.5281190196925736, + "grad_norm": 0.4239993393421173, + "learning_rate": 8.562902592189648e-05, + "loss": 0.06, + "step": 17110 + }, + { + "epoch": 0.5284276807210322, + "grad_norm": 0.3775506317615509, + "learning_rate": 8.560968620377921e-05, + "loss": 0.0476, + "step": 17120 + }, + { + "epoch": 0.5287363417494907, + "grad_norm": 0.40320679545402527, + "learning_rate": 8.559033566837951e-05, + "loss": 0.0541, + "step": 17130 + }, + { + "epoch": 0.5290450027779493, + "grad_norm": 0.3926146328449249, + "learning_rate": 8.557097432157551e-05, + "loss": 0.052, + "step": 17140 + }, + { + "epoch": 0.5293536638064078, + "grad_norm": 0.3471654951572418, + "learning_rate": 8.555160216924872e-05, + "loss": 0.0523, + "step": 17150 + }, + { + "epoch": 0.5296623248348663, + "grad_norm": 0.28206866979599, + "learning_rate": 8.55322192172839e-05, + "loss": 0.0499, + "step": 17160 + }, + { + "epoch": 0.5299709858633249, + "grad_norm": 0.37569937109947205, + "learning_rate": 8.551282547156902e-05, + "loss": 0.0526, + "step": 17170 + }, + { + "epoch": 0.5302796468917834, + "grad_norm": 0.344539999961853, + "learning_rate": 8.549342093799544e-05, + "loss": 0.0551, + "step": 17180 + }, + { + "epoch": 0.530588307920242, + "grad_norm": 0.4245755076408386, + "learning_rate": 8.547400562245773e-05, + "loss": 0.0514, + "step": 17190 + }, + { + "epoch": 0.5308969689487005, + "grad_norm": 0.9814355969429016, + "learning_rate": 8.545457953085374e-05, + "loss": 0.0519, + "step": 17200 + }, + { + "epoch": 0.531205629977159, + "grad_norm": 0.40632519125938416, + "learning_rate": 8.543514266908463e-05, + "loss": 0.0506, + "step": 17210 + }, + { + "epoch": 0.5315142910056176, + "grad_norm": 0.31123948097229004, + "learning_rate": 8.541569504305478e-05, + "loss": 0.0543, + "step": 17220 + }, + { + "epoch": 0.5318229520340761, + "grad_norm": 0.3500710427761078, + "learning_rate": 8.539623665867187e-05, + "loss": 0.0535, + "step": 17230 + }, + { + "epoch": 0.5321316130625348, + "grad_norm": 0.42944440245628357, + "learning_rate": 8.537676752184685e-05, + "loss": 0.0499, + "step": 17240 + }, + { + "epoch": 0.5324402740909933, + "grad_norm": 0.3957788646221161, + "learning_rate": 8.53572876384939e-05, + "loss": 0.0523, + "step": 17250 + }, + { + "epoch": 0.5327489351194519, + "grad_norm": 0.4571252763271332, + "learning_rate": 8.533779701453056e-05, + "loss": 0.055, + "step": 17260 + }, + { + "epoch": 0.5330575961479104, + "grad_norm": 0.7999482750892639, + "learning_rate": 8.53182956558775e-05, + "loss": 0.05, + "step": 17270 + }, + { + "epoch": 0.5333662571763689, + "grad_norm": 0.30643361806869507, + "learning_rate": 8.529878356845877e-05, + "loss": 0.052, + "step": 17280 + }, + { + "epoch": 0.5336749182048275, + "grad_norm": 0.290657103061676, + "learning_rate": 8.527926075820158e-05, + "loss": 0.0502, + "step": 17290 + }, + { + "epoch": 0.533983579233286, + "grad_norm": 0.2862887978553772, + "learning_rate": 8.525972723103648e-05, + "loss": 0.0507, + "step": 17300 + }, + { + "epoch": 0.5342922402617446, + "grad_norm": 0.3437122106552124, + "learning_rate": 8.524018299289722e-05, + "loss": 0.0536, + "step": 17310 + }, + { + "epoch": 0.5346009012902031, + "grad_norm": 0.3141625225543976, + "learning_rate": 8.522062804972083e-05, + "loss": 0.0541, + "step": 17320 + }, + { + "epoch": 0.5349095623186616, + "grad_norm": 0.4516737163066864, + "learning_rate": 8.520106240744759e-05, + "loss": 0.0521, + "step": 17330 + }, + { + "epoch": 0.5352182233471202, + "grad_norm": 0.4500333070755005, + "learning_rate": 8.518148607202102e-05, + "loss": 0.0466, + "step": 17340 + }, + { + "epoch": 0.5355268843755787, + "grad_norm": 0.505405843257904, + "learning_rate": 8.51618990493879e-05, + "loss": 0.0513, + "step": 17350 + }, + { + "epoch": 0.5358355454040373, + "grad_norm": 0.29994556307792664, + "learning_rate": 8.514230134549823e-05, + "loss": 0.0463, + "step": 17360 + }, + { + "epoch": 0.5361442064324958, + "grad_norm": 0.5562757253646851, + "learning_rate": 8.51226929663053e-05, + "loss": 0.0532, + "step": 17370 + }, + { + "epoch": 0.5364528674609543, + "grad_norm": 0.3303307592868805, + "learning_rate": 8.51030739177656e-05, + "loss": 0.0516, + "step": 17380 + }, + { + "epoch": 0.5367615284894129, + "grad_norm": 0.32225242257118225, + "learning_rate": 8.508344420583889e-05, + "loss": 0.0503, + "step": 17390 + }, + { + "epoch": 0.5370701895178714, + "grad_norm": 0.30680525302886963, + "learning_rate": 8.506380383648816e-05, + "loss": 0.0568, + "step": 17400 + }, + { + "epoch": 0.5373788505463301, + "grad_norm": 0.3343501091003418, + "learning_rate": 8.504415281567963e-05, + "loss": 0.0485, + "step": 17410 + }, + { + "epoch": 0.5376875115747886, + "grad_norm": 0.29994717240333557, + "learning_rate": 8.502449114938275e-05, + "loss": 0.0502, + "step": 17420 + }, + { + "epoch": 0.5379961726032472, + "grad_norm": 0.36205798387527466, + "learning_rate": 8.500481884357025e-05, + "loss": 0.0472, + "step": 17430 + }, + { + "epoch": 0.5383048336317057, + "grad_norm": 0.47732388973236084, + "learning_rate": 8.498513590421801e-05, + "loss": 0.0483, + "step": 17440 + }, + { + "epoch": 0.5386134946601642, + "grad_norm": 0.39158356189727783, + "learning_rate": 8.496544233730522e-05, + "loss": 0.0498, + "step": 17450 + }, + { + "epoch": 0.5389221556886228, + "grad_norm": 0.38831156492233276, + "learning_rate": 8.494573814881426e-05, + "loss": 0.0547, + "step": 17460 + }, + { + "epoch": 0.5392308167170813, + "grad_norm": 0.3954644799232483, + "learning_rate": 8.492602334473074e-05, + "loss": 0.0507, + "step": 17470 + }, + { + "epoch": 0.5395394777455399, + "grad_norm": 0.33294913172721863, + "learning_rate": 8.49062979310435e-05, + "loss": 0.049, + "step": 17480 + }, + { + "epoch": 0.5398481387739984, + "grad_norm": 0.42858296632766724, + "learning_rate": 8.488656191374458e-05, + "loss": 0.0518, + "step": 17490 + }, + { + "epoch": 0.5401567998024569, + "grad_norm": 0.3663235008716583, + "learning_rate": 8.48668152988293e-05, + "loss": 0.048, + "step": 17500 + }, + { + "epoch": 0.5404654608309155, + "grad_norm": 0.3770352900028229, + "learning_rate": 8.484705809229612e-05, + "loss": 0.0491, + "step": 17510 + }, + { + "epoch": 0.540774121859374, + "grad_norm": 0.4635190963745117, + "learning_rate": 8.482729030014677e-05, + "loss": 0.0524, + "step": 17520 + }, + { + "epoch": 0.5410827828878326, + "grad_norm": 0.36072149872779846, + "learning_rate": 8.48075119283862e-05, + "loss": 0.0527, + "step": 17530 + }, + { + "epoch": 0.5413914439162911, + "grad_norm": 0.33703434467315674, + "learning_rate": 8.478772298302254e-05, + "loss": 0.0544, + "step": 17540 + }, + { + "epoch": 0.5417001049447496, + "grad_norm": 0.3327986001968384, + "learning_rate": 8.476792347006716e-05, + "loss": 0.0529, + "step": 17550 + }, + { + "epoch": 0.5420087659732082, + "grad_norm": 0.32773876190185547, + "learning_rate": 8.474811339553462e-05, + "loss": 0.0516, + "step": 17560 + }, + { + "epoch": 0.5423174270016667, + "grad_norm": 0.4089835584163666, + "learning_rate": 8.47282927654427e-05, + "loss": 0.054, + "step": 17570 + }, + { + "epoch": 0.5426260880301254, + "grad_norm": 0.4062270522117615, + "learning_rate": 8.470846158581238e-05, + "loss": 0.0536, + "step": 17580 + }, + { + "epoch": 0.5429347490585839, + "grad_norm": 0.5677525401115417, + "learning_rate": 8.468861986266787e-05, + "loss": 0.0538, + "step": 17590 + }, + { + "epoch": 0.5432434100870424, + "grad_norm": 0.3811545968055725, + "learning_rate": 8.466876760203654e-05, + "loss": 0.0536, + "step": 17600 + }, + { + "epoch": 0.543552071115501, + "grad_norm": 0.3321947753429413, + "learning_rate": 8.464890480994898e-05, + "loss": 0.0526, + "step": 17610 + }, + { + "epoch": 0.5438607321439595, + "grad_norm": 0.33143332600593567, + "learning_rate": 8.462903149243899e-05, + "loss": 0.0476, + "step": 17620 + }, + { + "epoch": 0.5441693931724181, + "grad_norm": 0.3068641126155853, + "learning_rate": 8.460914765554357e-05, + "loss": 0.0585, + "step": 17630 + }, + { + "epoch": 0.5444780542008766, + "grad_norm": 0.595100998878479, + "learning_rate": 8.458925330530288e-05, + "loss": 0.0523, + "step": 17640 + }, + { + "epoch": 0.5447867152293352, + "grad_norm": 0.3672502636909485, + "learning_rate": 8.456934844776032e-05, + "loss": 0.0534, + "step": 17650 + }, + { + "epoch": 0.5450953762577937, + "grad_norm": 0.745363175868988, + "learning_rate": 8.454943308896246e-05, + "loss": 0.055, + "step": 17660 + }, + { + "epoch": 0.5454040372862522, + "grad_norm": 0.3409796953201294, + "learning_rate": 8.452950723495905e-05, + "loss": 0.0482, + "step": 17670 + }, + { + "epoch": 0.5457126983147108, + "grad_norm": 0.4061517119407654, + "learning_rate": 8.450957089180303e-05, + "loss": 0.0496, + "step": 17680 + }, + { + "epoch": 0.5460213593431693, + "grad_norm": 0.3445150852203369, + "learning_rate": 8.448962406555055e-05, + "loss": 0.0555, + "step": 17690 + }, + { + "epoch": 0.5463300203716279, + "grad_norm": 0.2962495982646942, + "learning_rate": 8.446966676226093e-05, + "loss": 0.0486, + "step": 17700 + }, + { + "epoch": 0.5466386814000864, + "grad_norm": 0.4515039026737213, + "learning_rate": 8.444969898799667e-05, + "loss": 0.0488, + "step": 17710 + }, + { + "epoch": 0.5469473424285449, + "grad_norm": 0.3441462516784668, + "learning_rate": 8.442972074882343e-05, + "loss": 0.0489, + "step": 17720 + }, + { + "epoch": 0.5472560034570035, + "grad_norm": 0.3469010293483734, + "learning_rate": 8.44097320508101e-05, + "loss": 0.0515, + "step": 17730 + }, + { + "epoch": 0.547564664485462, + "grad_norm": 0.32305458188056946, + "learning_rate": 8.43897329000287e-05, + "loss": 0.0466, + "step": 17740 + }, + { + "epoch": 0.5478733255139207, + "grad_norm": 0.3854775130748749, + "learning_rate": 8.436972330255448e-05, + "loss": 0.0511, + "step": 17750 + }, + { + "epoch": 0.5481819865423792, + "grad_norm": 0.29163286089897156, + "learning_rate": 8.434970326446579e-05, + "loss": 0.0449, + "step": 17760 + }, + { + "epoch": 0.5484906475708377, + "grad_norm": 0.272288054227829, + "learning_rate": 8.432967279184418e-05, + "loss": 0.0502, + "step": 17770 + }, + { + "epoch": 0.5487993085992963, + "grad_norm": 0.2911209762096405, + "learning_rate": 8.430963189077441e-05, + "loss": 0.0472, + "step": 17780 + }, + { + "epoch": 0.5491079696277548, + "grad_norm": 0.42177659273147583, + "learning_rate": 8.428958056734437e-05, + "loss": 0.0542, + "step": 17790 + }, + { + "epoch": 0.5494166306562134, + "grad_norm": 0.40070247650146484, + "learning_rate": 8.426951882764513e-05, + "loss": 0.0492, + "step": 17800 + }, + { + "epoch": 0.5497252916846719, + "grad_norm": 0.3183360993862152, + "learning_rate": 8.424944667777089e-05, + "loss": 0.0505, + "step": 17810 + }, + { + "epoch": 0.5500339527131305, + "grad_norm": 0.291474312543869, + "learning_rate": 8.422936412381905e-05, + "loss": 0.0562, + "step": 17820 + }, + { + "epoch": 0.550342613741589, + "grad_norm": 0.3124450743198395, + "learning_rate": 8.420927117189017e-05, + "loss": 0.0473, + "step": 17830 + }, + { + "epoch": 0.5506512747700475, + "grad_norm": 0.45204585790634155, + "learning_rate": 8.418916782808795e-05, + "loss": 0.0556, + "step": 17840 + }, + { + "epoch": 0.5509599357985061, + "grad_norm": 0.3298812806606293, + "learning_rate": 8.416905409851926e-05, + "loss": 0.0538, + "step": 17850 + }, + { + "epoch": 0.5512685968269646, + "grad_norm": 0.3093234598636627, + "learning_rate": 8.41489299892941e-05, + "loss": 0.0508, + "step": 17860 + }, + { + "epoch": 0.5515772578554232, + "grad_norm": 0.7432244420051575, + "learning_rate": 8.412879550652566e-05, + "loss": 0.0534, + "step": 17870 + }, + { + "epoch": 0.5518859188838817, + "grad_norm": 0.36565476655960083, + "learning_rate": 8.410865065633029e-05, + "loss": 0.0474, + "step": 17880 + }, + { + "epoch": 0.5521945799123402, + "grad_norm": 0.3032131493091583, + "learning_rate": 8.408849544482742e-05, + "loss": 0.0494, + "step": 17890 + }, + { + "epoch": 0.5525032409407988, + "grad_norm": 0.30438119173049927, + "learning_rate": 8.406832987813968e-05, + "loss": 0.0489, + "step": 17900 + }, + { + "epoch": 0.5528119019692573, + "grad_norm": 0.3080512583255768, + "learning_rate": 8.404815396239286e-05, + "loss": 0.0514, + "step": 17910 + }, + { + "epoch": 0.553120562997716, + "grad_norm": 0.706312894821167, + "learning_rate": 8.402796770371587e-05, + "loss": 0.056, + "step": 17920 + }, + { + "epoch": 0.5534292240261744, + "grad_norm": 0.3443800210952759, + "learning_rate": 8.400777110824071e-05, + "loss": 0.0516, + "step": 17930 + }, + { + "epoch": 0.553737885054633, + "grad_norm": 0.2736016809940338, + "learning_rate": 8.398756418210263e-05, + "loss": 0.0516, + "step": 17940 + }, + { + "epoch": 0.5540465460830916, + "grad_norm": 0.2821025252342224, + "learning_rate": 8.396734693143993e-05, + "loss": 0.0506, + "step": 17950 + }, + { + "epoch": 0.5543552071115501, + "grad_norm": 0.349929541349411, + "learning_rate": 8.39471193623941e-05, + "loss": 0.0534, + "step": 17960 + }, + { + "epoch": 0.5546638681400087, + "grad_norm": 0.2711075246334076, + "learning_rate": 8.392688148110974e-05, + "loss": 0.0515, + "step": 17970 + }, + { + "epoch": 0.5549725291684672, + "grad_norm": 0.34008845686912537, + "learning_rate": 8.390663329373456e-05, + "loss": 0.0467, + "step": 17980 + }, + { + "epoch": 0.5552811901969258, + "grad_norm": 0.3312731087207794, + "learning_rate": 8.388637480641944e-05, + "loss": 0.0497, + "step": 17990 + }, + { + "epoch": 0.5555898512253843, + "grad_norm": 0.3364715576171875, + "learning_rate": 8.386610602531837e-05, + "loss": 0.0501, + "step": 18000 + }, + { + "epoch": 0.5558985122538428, + "grad_norm": 0.32990071177482605, + "learning_rate": 8.384582695658847e-05, + "loss": 0.0501, + "step": 18010 + }, + { + "epoch": 0.5562071732823014, + "grad_norm": 0.3851384222507477, + "learning_rate": 8.382553760638999e-05, + "loss": 0.0536, + "step": 18020 + }, + { + "epoch": 0.5565158343107599, + "grad_norm": 0.3007829785346985, + "learning_rate": 8.380523798088631e-05, + "loss": 0.0489, + "step": 18030 + }, + { + "epoch": 0.5568244953392185, + "grad_norm": 0.37661704421043396, + "learning_rate": 8.378492808624389e-05, + "loss": 0.053, + "step": 18040 + }, + { + "epoch": 0.557133156367677, + "grad_norm": 0.3465919494628906, + "learning_rate": 8.376460792863237e-05, + "loss": 0.0534, + "step": 18050 + }, + { + "epoch": 0.5574418173961355, + "grad_norm": 0.4007294774055481, + "learning_rate": 8.374427751422444e-05, + "loss": 0.0483, + "step": 18060 + }, + { + "epoch": 0.5577504784245941, + "grad_norm": 0.6347354650497437, + "learning_rate": 8.3723936849196e-05, + "loss": 0.0538, + "step": 18070 + }, + { + "epoch": 0.5580591394530526, + "grad_norm": 0.4231362044811249, + "learning_rate": 8.370358593972595e-05, + "loss": 0.0516, + "step": 18080 + }, + { + "epoch": 0.5583678004815112, + "grad_norm": 0.3741329610347748, + "learning_rate": 8.36832247919964e-05, + "loss": 0.0517, + "step": 18090 + }, + { + "epoch": 0.5586764615099697, + "grad_norm": 0.25527191162109375, + "learning_rate": 8.36628534121925e-05, + "loss": 0.0476, + "step": 18100 + }, + { + "epoch": 0.5589851225384282, + "grad_norm": 0.36410582065582275, + "learning_rate": 8.364247180650254e-05, + "loss": 0.0538, + "step": 18110 + }, + { + "epoch": 0.5592937835668869, + "grad_norm": 0.2928015887737274, + "learning_rate": 8.362207998111794e-05, + "loss": 0.0532, + "step": 18120 + }, + { + "epoch": 0.5596024445953454, + "grad_norm": 0.2802152633666992, + "learning_rate": 8.360167794223318e-05, + "loss": 0.0459, + "step": 18130 + }, + { + "epoch": 0.559911105623804, + "grad_norm": 0.43582436442375183, + "learning_rate": 8.358126569604586e-05, + "loss": 0.0605, + "step": 18140 + }, + { + "epoch": 0.5602197666522625, + "grad_norm": 0.3007802665233612, + "learning_rate": 8.356084324875668e-05, + "loss": 0.0488, + "step": 18150 + }, + { + "epoch": 0.5605284276807211, + "grad_norm": 0.33904072642326355, + "learning_rate": 8.354041060656945e-05, + "loss": 0.0505, + "step": 18160 + }, + { + "epoch": 0.5608370887091796, + "grad_norm": 0.4120822548866272, + "learning_rate": 8.351996777569106e-05, + "loss": 0.0544, + "step": 18170 + }, + { + "epoch": 0.5611457497376381, + "grad_norm": 0.37621235847473145, + "learning_rate": 8.349951476233148e-05, + "loss": 0.0536, + "step": 18180 + }, + { + "epoch": 0.5614544107660967, + "grad_norm": 0.40492427349090576, + "learning_rate": 8.347905157270386e-05, + "loss": 0.0512, + "step": 18190 + }, + { + "epoch": 0.5617630717945552, + "grad_norm": 0.4069778025150299, + "learning_rate": 8.345857821302432e-05, + "loss": 0.0528, + "step": 18200 + }, + { + "epoch": 0.5620717328230138, + "grad_norm": 0.334403932094574, + "learning_rate": 8.343809468951213e-05, + "loss": 0.0487, + "step": 18210 + }, + { + "epoch": 0.5623803938514723, + "grad_norm": 0.2925446927547455, + "learning_rate": 8.341760100838965e-05, + "loss": 0.0566, + "step": 18220 + }, + { + "epoch": 0.5626890548799308, + "grad_norm": 0.38633427023887634, + "learning_rate": 8.339709717588233e-05, + "loss": 0.0498, + "step": 18230 + }, + { + "epoch": 0.5629977159083894, + "grad_norm": 0.3496672213077545, + "learning_rate": 8.33765831982187e-05, + "loss": 0.0517, + "step": 18240 + }, + { + "epoch": 0.5633063769368479, + "grad_norm": 0.42831331491470337, + "learning_rate": 8.335605908163035e-05, + "loss": 0.0509, + "step": 18250 + }, + { + "epoch": 0.5636150379653065, + "grad_norm": 0.4191705882549286, + "learning_rate": 8.333552483235196e-05, + "loss": 0.0499, + "step": 18260 + }, + { + "epoch": 0.563923698993765, + "grad_norm": 0.3300780951976776, + "learning_rate": 8.33149804566213e-05, + "loss": 0.0523, + "step": 18270 + }, + { + "epoch": 0.5642323600222235, + "grad_norm": 0.32811686396598816, + "learning_rate": 8.329442596067921e-05, + "loss": 0.05, + "step": 18280 + }, + { + "epoch": 0.5645410210506822, + "grad_norm": 0.5326352119445801, + "learning_rate": 8.32738613507696e-05, + "loss": 0.0545, + "step": 18290 + }, + { + "epoch": 0.5648496820791407, + "grad_norm": 0.36862796545028687, + "learning_rate": 8.325328663313946e-05, + "loss": 0.0488, + "step": 18300 + }, + { + "epoch": 0.5651583431075993, + "grad_norm": 0.4030299186706543, + "learning_rate": 8.323270181403884e-05, + "loss": 0.0532, + "step": 18310 + }, + { + "epoch": 0.5654670041360578, + "grad_norm": 0.32199105620384216, + "learning_rate": 8.321210689972086e-05, + "loss": 0.0474, + "step": 18320 + }, + { + "epoch": 0.5657756651645164, + "grad_norm": 0.29485368728637695, + "learning_rate": 8.319150189644174e-05, + "loss": 0.0533, + "step": 18330 + }, + { + "epoch": 0.5660843261929749, + "grad_norm": 0.42078888416290283, + "learning_rate": 8.31708868104607e-05, + "loss": 0.0496, + "step": 18340 + }, + { + "epoch": 0.5663929872214334, + "grad_norm": 0.2808919847011566, + "learning_rate": 8.315026164804007e-05, + "loss": 0.0496, + "step": 18350 + }, + { + "epoch": 0.566701648249892, + "grad_norm": 0.37500128149986267, + "learning_rate": 8.312962641544524e-05, + "loss": 0.05, + "step": 18360 + }, + { + "epoch": 0.5670103092783505, + "grad_norm": 0.27475330233573914, + "learning_rate": 8.310898111894465e-05, + "loss": 0.0551, + "step": 18370 + }, + { + "epoch": 0.5673189703068091, + "grad_norm": 0.3902367949485779, + "learning_rate": 8.308832576480977e-05, + "loss": 0.0581, + "step": 18380 + }, + { + "epoch": 0.5676276313352676, + "grad_norm": 0.3298451900482178, + "learning_rate": 8.306766035931519e-05, + "loss": 0.0498, + "step": 18390 + }, + { + "epoch": 0.5679362923637261, + "grad_norm": 0.36443620920181274, + "learning_rate": 8.304698490873847e-05, + "loss": 0.0535, + "step": 18400 + }, + { + "epoch": 0.5682449533921847, + "grad_norm": 0.2968038022518158, + "learning_rate": 8.30262994193603e-05, + "loss": 0.0534, + "step": 18410 + }, + { + "epoch": 0.5685536144206432, + "grad_norm": 0.37626656889915466, + "learning_rate": 8.300560389746438e-05, + "loss": 0.0502, + "step": 18420 + }, + { + "epoch": 0.5688622754491018, + "grad_norm": 0.2970075309276581, + "learning_rate": 8.298489834933745e-05, + "loss": 0.0525, + "step": 18430 + }, + { + "epoch": 0.5691709364775603, + "grad_norm": 0.35683152079582214, + "learning_rate": 8.296418278126934e-05, + "loss": 0.0475, + "step": 18440 + }, + { + "epoch": 0.5694795975060188, + "grad_norm": 0.41440099477767944, + "learning_rate": 8.294345719955284e-05, + "loss": 0.0534, + "step": 18450 + }, + { + "epoch": 0.5697882585344775, + "grad_norm": 0.30822646617889404, + "learning_rate": 8.29227216104839e-05, + "loss": 0.0534, + "step": 18460 + }, + { + "epoch": 0.570096919562936, + "grad_norm": 0.4136391282081604, + "learning_rate": 8.290197602036137e-05, + "loss": 0.0485, + "step": 18470 + }, + { + "epoch": 0.5704055805913946, + "grad_norm": 0.2926332950592041, + "learning_rate": 8.288122043548725e-05, + "loss": 0.0514, + "step": 18480 + }, + { + "epoch": 0.5707142416198531, + "grad_norm": 0.28662338852882385, + "learning_rate": 8.286045486216657e-05, + "loss": 0.0506, + "step": 18490 + }, + { + "epoch": 0.5710229026483116, + "grad_norm": 0.24920538067817688, + "learning_rate": 8.283967930670733e-05, + "loss": 0.0498, + "step": 18500 + }, + { + "epoch": 0.5713315636767702, + "grad_norm": 0.39629659056663513, + "learning_rate": 8.281889377542058e-05, + "loss": 0.0531, + "step": 18510 + }, + { + "epoch": 0.5716402247052287, + "grad_norm": 0.3787921369075775, + "learning_rate": 8.279809827462045e-05, + "loss": 0.0512, + "step": 18520 + }, + { + "epoch": 0.5719488857336873, + "grad_norm": 0.336675763130188, + "learning_rate": 8.277729281062402e-05, + "loss": 0.0485, + "step": 18530 + }, + { + "epoch": 0.5722575467621458, + "grad_norm": 0.3886491060256958, + "learning_rate": 8.27564773897515e-05, + "loss": 0.0548, + "step": 18540 + }, + { + "epoch": 0.5725662077906044, + "grad_norm": 0.4887841045856476, + "learning_rate": 8.273565201832602e-05, + "loss": 0.0514, + "step": 18550 + }, + { + "epoch": 0.5728748688190629, + "grad_norm": 0.2538938820362091, + "learning_rate": 8.27148167026738e-05, + "loss": 0.0491, + "step": 18560 + }, + { + "epoch": 0.5731835298475214, + "grad_norm": 0.3732413649559021, + "learning_rate": 8.269397144912405e-05, + "loss": 0.0548, + "step": 18570 + }, + { + "epoch": 0.57349219087598, + "grad_norm": 0.3354870676994324, + "learning_rate": 8.267311626400899e-05, + "loss": 0.0476, + "step": 18580 + }, + { + "epoch": 0.5738008519044385, + "grad_norm": 0.4127795398235321, + "learning_rate": 8.26522511536639e-05, + "loss": 0.053, + "step": 18590 + }, + { + "epoch": 0.5741095129328971, + "grad_norm": 0.3563206195831299, + "learning_rate": 8.263137612442706e-05, + "loss": 0.05, + "step": 18600 + }, + { + "epoch": 0.5744181739613556, + "grad_norm": 0.2713235318660736, + "learning_rate": 8.261049118263971e-05, + "loss": 0.0523, + "step": 18610 + }, + { + "epoch": 0.5747268349898141, + "grad_norm": 0.36671093106269836, + "learning_rate": 8.258959633464619e-05, + "loss": 0.0505, + "step": 18620 + }, + { + "epoch": 0.5750354960182728, + "grad_norm": 0.3954453468322754, + "learning_rate": 8.256869158679377e-05, + "loss": 0.0502, + "step": 18630 + }, + { + "epoch": 0.5753441570467313, + "grad_norm": 0.6135392785072327, + "learning_rate": 8.254777694543278e-05, + "loss": 0.0498, + "step": 18640 + }, + { + "epoch": 0.5756528180751899, + "grad_norm": 0.3863522410392761, + "learning_rate": 8.252685241691651e-05, + "loss": 0.0441, + "step": 18650 + }, + { + "epoch": 0.5759614791036484, + "grad_norm": 0.37435486912727356, + "learning_rate": 8.250591800760133e-05, + "loss": 0.0528, + "step": 18660 + }, + { + "epoch": 0.5762701401321069, + "grad_norm": 0.3249048590660095, + "learning_rate": 8.248497372384649e-05, + "loss": 0.045, + "step": 18670 + }, + { + "epoch": 0.5765788011605655, + "grad_norm": 0.7181516885757446, + "learning_rate": 8.246401957201437e-05, + "loss": 0.0512, + "step": 18680 + }, + { + "epoch": 0.576887462189024, + "grad_norm": 0.3018963932991028, + "learning_rate": 8.244305555847027e-05, + "loss": 0.0523, + "step": 18690 + }, + { + "epoch": 0.5771961232174826, + "grad_norm": 0.5235023498535156, + "learning_rate": 8.24220816895825e-05, + "loss": 0.0504, + "step": 18700 + }, + { + "epoch": 0.5775047842459411, + "grad_norm": 0.3321242928504944, + "learning_rate": 8.240109797172237e-05, + "loss": 0.0509, + "step": 18710 + }, + { + "epoch": 0.5778134452743997, + "grad_norm": 0.32210543751716614, + "learning_rate": 8.238010441126416e-05, + "loss": 0.0529, + "step": 18720 + }, + { + "epoch": 0.5781221063028582, + "grad_norm": 0.40363243222236633, + "learning_rate": 8.23591010145852e-05, + "loss": 0.0443, + "step": 18730 + }, + { + "epoch": 0.5784307673313167, + "grad_norm": 0.2838019132614136, + "learning_rate": 8.233808778806571e-05, + "loss": 0.0551, + "step": 18740 + }, + { + "epoch": 0.5787394283597753, + "grad_norm": 0.4363957345485687, + "learning_rate": 8.231706473808903e-05, + "loss": 0.0503, + "step": 18750 + }, + { + "epoch": 0.5790480893882338, + "grad_norm": 0.6531949639320374, + "learning_rate": 8.229603187104133e-05, + "loss": 0.0484, + "step": 18760 + }, + { + "epoch": 0.5793567504166924, + "grad_norm": 0.5304802060127258, + "learning_rate": 8.22749891933119e-05, + "loss": 0.0515, + "step": 18770 + }, + { + "epoch": 0.5796654114451509, + "grad_norm": 0.2790489196777344, + "learning_rate": 8.225393671129291e-05, + "loss": 0.0529, + "step": 18780 + }, + { + "epoch": 0.5799740724736094, + "grad_norm": 0.3958154022693634, + "learning_rate": 8.223287443137957e-05, + "loss": 0.0481, + "step": 18790 + }, + { + "epoch": 0.580282733502068, + "grad_norm": 0.3212132453918457, + "learning_rate": 8.221180235997004e-05, + "loss": 0.0533, + "step": 18800 + }, + { + "epoch": 0.5805913945305265, + "grad_norm": 0.2867889702320099, + "learning_rate": 8.219072050346544e-05, + "loss": 0.0471, + "step": 18810 + }, + { + "epoch": 0.5809000555589852, + "grad_norm": 0.31859657168388367, + "learning_rate": 8.216962886826992e-05, + "loss": 0.0528, + "step": 18820 + }, + { + "epoch": 0.5812087165874437, + "grad_norm": 0.31635782122612, + "learning_rate": 8.214852746079054e-05, + "loss": 0.0504, + "step": 18830 + }, + { + "epoch": 0.5815173776159022, + "grad_norm": 0.2788955569267273, + "learning_rate": 8.212741628743732e-05, + "loss": 0.0521, + "step": 18840 + }, + { + "epoch": 0.5818260386443608, + "grad_norm": 0.2732284665107727, + "learning_rate": 8.210629535462333e-05, + "loss": 0.0515, + "step": 18850 + }, + { + "epoch": 0.5821346996728193, + "grad_norm": 0.2941543757915497, + "learning_rate": 8.208516466876453e-05, + "loss": 0.052, + "step": 18860 + }, + { + "epoch": 0.5824433607012779, + "grad_norm": 0.2913353741168976, + "learning_rate": 8.206402423627986e-05, + "loss": 0.0479, + "step": 18870 + }, + { + "epoch": 0.5827520217297364, + "grad_norm": 0.4601471722126007, + "learning_rate": 8.204287406359124e-05, + "loss": 0.051, + "step": 18880 + }, + { + "epoch": 0.583060682758195, + "grad_norm": 0.3795967102050781, + "learning_rate": 8.20217141571235e-05, + "loss": 0.0507, + "step": 18890 + }, + { + "epoch": 0.5833693437866535, + "grad_norm": 0.3833085298538208, + "learning_rate": 8.200054452330449e-05, + "loss": 0.0489, + "step": 18900 + }, + { + "epoch": 0.583678004815112, + "grad_norm": 0.28051459789276123, + "learning_rate": 8.197936516856499e-05, + "loss": 0.0488, + "step": 18910 + }, + { + "epoch": 0.5839866658435706, + "grad_norm": 0.3608768880367279, + "learning_rate": 8.195817609933871e-05, + "loss": 0.0545, + "step": 18920 + }, + { + "epoch": 0.5842953268720291, + "grad_norm": 0.3614247441291809, + "learning_rate": 8.193697732206233e-05, + "loss": 0.0513, + "step": 18930 + }, + { + "epoch": 0.5846039879004877, + "grad_norm": 0.37097135186195374, + "learning_rate": 8.19157688431755e-05, + "loss": 0.0462, + "step": 18940 + }, + { + "epoch": 0.5849126489289462, + "grad_norm": 0.2305632382631302, + "learning_rate": 8.189455066912077e-05, + "loss": 0.0482, + "step": 18950 + }, + { + "epoch": 0.5852213099574047, + "grad_norm": 0.2970634698867798, + "learning_rate": 8.187332280634369e-05, + "loss": 0.0467, + "step": 18960 + }, + { + "epoch": 0.5855299709858633, + "grad_norm": 0.276040643453598, + "learning_rate": 8.18520852612927e-05, + "loss": 0.0501, + "step": 18970 + }, + { + "epoch": 0.5858386320143218, + "grad_norm": 0.3468988835811615, + "learning_rate": 8.183083804041921e-05, + "loss": 0.0529, + "step": 18980 + }, + { + "epoch": 0.5861472930427805, + "grad_norm": 0.33790600299835205, + "learning_rate": 8.180958115017757e-05, + "loss": 0.0514, + "step": 18990 + }, + { + "epoch": 0.586455954071239, + "grad_norm": 0.5833762884140015, + "learning_rate": 8.178831459702505e-05, + "loss": 0.0482, + "step": 19000 + }, + { + "epoch": 0.5867646150996975, + "grad_norm": 0.28234949707984924, + "learning_rate": 8.17670383874219e-05, + "loss": 0.0484, + "step": 19010 + }, + { + "epoch": 0.5870732761281561, + "grad_norm": 0.3372311294078827, + "learning_rate": 8.174575252783124e-05, + "loss": 0.0508, + "step": 19020 + }, + { + "epoch": 0.5873819371566146, + "grad_norm": 0.5835558176040649, + "learning_rate": 8.172445702471914e-05, + "loss": 0.0512, + "step": 19030 + }, + { + "epoch": 0.5876905981850732, + "grad_norm": 0.4186050295829773, + "learning_rate": 8.170315188455466e-05, + "loss": 0.0549, + "step": 19040 + }, + { + "epoch": 0.5879992592135317, + "grad_norm": 0.3783319890499115, + "learning_rate": 8.168183711380969e-05, + "loss": 0.0492, + "step": 19050 + }, + { + "epoch": 0.5883079202419903, + "grad_norm": 0.41992297768592834, + "learning_rate": 8.166051271895913e-05, + "loss": 0.0554, + "step": 19060 + }, + { + "epoch": 0.5886165812704488, + "grad_norm": 0.6553993225097656, + "learning_rate": 8.163917870648075e-05, + "loss": 0.0545, + "step": 19070 + }, + { + "epoch": 0.5889252422989073, + "grad_norm": 0.36333346366882324, + "learning_rate": 8.161783508285526e-05, + "loss": 0.0528, + "step": 19080 + }, + { + "epoch": 0.5892339033273659, + "grad_norm": 0.5036302208900452, + "learning_rate": 8.159648185456628e-05, + "loss": 0.049, + "step": 19090 + }, + { + "epoch": 0.5895425643558244, + "grad_norm": 0.3632667064666748, + "learning_rate": 8.157511902810038e-05, + "loss": 0.0547, + "step": 19100 + }, + { + "epoch": 0.589851225384283, + "grad_norm": 0.5241316556930542, + "learning_rate": 8.155374660994701e-05, + "loss": 0.0556, + "step": 19110 + }, + { + "epoch": 0.5901598864127415, + "grad_norm": 0.3806178569793701, + "learning_rate": 8.153236460659857e-05, + "loss": 0.0536, + "step": 19120 + }, + { + "epoch": 0.5904685474412, + "grad_norm": 0.3719522953033447, + "learning_rate": 8.151097302455031e-05, + "loss": 0.0516, + "step": 19130 + }, + { + "epoch": 0.5907772084696586, + "grad_norm": 0.4219362139701843, + "learning_rate": 8.148957187030044e-05, + "loss": 0.0507, + "step": 19140 + }, + { + "epoch": 0.5910858694981171, + "grad_norm": 0.3647024631500244, + "learning_rate": 8.146816115035006e-05, + "loss": 0.0503, + "step": 19150 + }, + { + "epoch": 0.5913945305265758, + "grad_norm": 0.34472277760505676, + "learning_rate": 8.14467408712032e-05, + "loss": 0.0492, + "step": 19160 + }, + { + "epoch": 0.5917031915550343, + "grad_norm": 0.3994935154914856, + "learning_rate": 8.142531103936678e-05, + "loss": 0.0465, + "step": 19170 + }, + { + "epoch": 0.5920118525834928, + "grad_norm": 0.3359880745410919, + "learning_rate": 8.14038716613506e-05, + "loss": 0.0541, + "step": 19180 + }, + { + "epoch": 0.5923205136119514, + "grad_norm": 0.3544779419898987, + "learning_rate": 8.138242274366736e-05, + "loss": 0.0509, + "step": 19190 + }, + { + "epoch": 0.5926291746404099, + "grad_norm": 0.3036300837993622, + "learning_rate": 8.136096429283271e-05, + "loss": 0.0529, + "step": 19200 + }, + { + "epoch": 0.5929378356688685, + "grad_norm": 0.3530232906341553, + "learning_rate": 8.133949631536515e-05, + "loss": 0.0463, + "step": 19210 + }, + { + "epoch": 0.593246496697327, + "grad_norm": 0.33874577283859253, + "learning_rate": 8.131801881778607e-05, + "loss": 0.0518, + "step": 19220 + }, + { + "epoch": 0.5935551577257855, + "grad_norm": 0.2964797914028168, + "learning_rate": 8.129653180661978e-05, + "loss": 0.0515, + "step": 19230 + }, + { + "epoch": 0.5938638187542441, + "grad_norm": 0.35379430651664734, + "learning_rate": 8.127503528839346e-05, + "loss": 0.0511, + "step": 19240 + }, + { + "epoch": 0.5941724797827026, + "grad_norm": 0.493857204914093, + "learning_rate": 8.125352926963721e-05, + "loss": 0.054, + "step": 19250 + }, + { + "epoch": 0.5944811408111612, + "grad_norm": 0.44851094484329224, + "learning_rate": 8.123201375688395e-05, + "loss": 0.0533, + "step": 19260 + }, + { + "epoch": 0.5947898018396197, + "grad_norm": 0.2705232501029968, + "learning_rate": 8.121048875666954e-05, + "loss": 0.0436, + "step": 19270 + }, + { + "epoch": 0.5950984628680783, + "grad_norm": 0.37455153465270996, + "learning_rate": 8.118895427553274e-05, + "loss": 0.058, + "step": 19280 + }, + { + "epoch": 0.5954071238965368, + "grad_norm": 0.5712986588478088, + "learning_rate": 8.116741032001511e-05, + "loss": 0.0515, + "step": 19290 + }, + { + "epoch": 0.5957157849249953, + "grad_norm": 0.36130911111831665, + "learning_rate": 8.114585689666114e-05, + "loss": 0.053, + "step": 19300 + }, + { + "epoch": 0.5960244459534539, + "grad_norm": 0.28279992938041687, + "learning_rate": 8.112429401201821e-05, + "loss": 0.0464, + "step": 19310 + }, + { + "epoch": 0.5963331069819124, + "grad_norm": 0.3452304005622864, + "learning_rate": 8.110272167263656e-05, + "loss": 0.0531, + "step": 19320 + }, + { + "epoch": 0.596641768010371, + "grad_norm": 0.4102878272533417, + "learning_rate": 8.108113988506929e-05, + "loss": 0.0489, + "step": 19330 + }, + { + "epoch": 0.5969504290388296, + "grad_norm": 0.35880786180496216, + "learning_rate": 8.105954865587235e-05, + "loss": 0.051, + "step": 19340 + }, + { + "epoch": 0.5972590900672881, + "grad_norm": 0.3711908161640167, + "learning_rate": 8.103794799160463e-05, + "loss": 0.0481, + "step": 19350 + }, + { + "epoch": 0.5975677510957467, + "grad_norm": 0.2806364893913269, + "learning_rate": 8.101633789882781e-05, + "loss": 0.0471, + "step": 19360 + }, + { + "epoch": 0.5978764121242052, + "grad_norm": 0.27733314037323, + "learning_rate": 8.099471838410648e-05, + "loss": 0.052, + "step": 19370 + }, + { + "epoch": 0.5981850731526638, + "grad_norm": 0.3077859878540039, + "learning_rate": 8.097308945400806e-05, + "loss": 0.0484, + "step": 19380 + }, + { + "epoch": 0.5984937341811223, + "grad_norm": 0.34305983781814575, + "learning_rate": 8.095145111510288e-05, + "loss": 0.0495, + "step": 19390 + }, + { + "epoch": 0.5988023952095808, + "grad_norm": 0.28174102306365967, + "learning_rate": 8.092980337396406e-05, + "loss": 0.0499, + "step": 19400 + }, + { + "epoch": 0.5991110562380394, + "grad_norm": 0.30206531286239624, + "learning_rate": 8.090814623716763e-05, + "loss": 0.0494, + "step": 19410 + }, + { + "epoch": 0.5994197172664979, + "grad_norm": 0.3351582884788513, + "learning_rate": 8.088647971129246e-05, + "loss": 0.0485, + "step": 19420 + }, + { + "epoch": 0.5997283782949565, + "grad_norm": 0.43761518597602844, + "learning_rate": 8.086480380292026e-05, + "loss": 0.0485, + "step": 19430 + }, + { + "epoch": 0.600037039323415, + "grad_norm": 0.4074511229991913, + "learning_rate": 8.084311851863562e-05, + "loss": 0.0516, + "step": 19440 + }, + { + "epoch": 0.6003457003518736, + "grad_norm": 0.4073620140552521, + "learning_rate": 8.082142386502591e-05, + "loss": 0.0458, + "step": 19450 + }, + { + "epoch": 0.6006543613803321, + "grad_norm": 0.42395099997520447, + "learning_rate": 8.079971984868145e-05, + "loss": 0.0487, + "step": 19460 + }, + { + "epoch": 0.6009630224087906, + "grad_norm": 0.3064388036727905, + "learning_rate": 8.077800647619532e-05, + "loss": 0.0471, + "step": 19470 + }, + { + "epoch": 0.6012716834372492, + "grad_norm": 0.32255256175994873, + "learning_rate": 8.075628375416345e-05, + "loss": 0.0493, + "step": 19480 + }, + { + "epoch": 0.6015803444657077, + "grad_norm": 0.320661723613739, + "learning_rate": 8.073455168918464e-05, + "loss": 0.049, + "step": 19490 + }, + { + "epoch": 0.6018890054941664, + "grad_norm": 0.3809105455875397, + "learning_rate": 8.071281028786055e-05, + "loss": 0.0489, + "step": 19500 + }, + { + "epoch": 0.6021976665226249, + "grad_norm": 0.5274997353553772, + "learning_rate": 8.069105955679562e-05, + "loss": 0.0558, + "step": 19510 + }, + { + "epoch": 0.6025063275510834, + "grad_norm": 0.3363320231437683, + "learning_rate": 8.066929950259713e-05, + "loss": 0.0486, + "step": 19520 + }, + { + "epoch": 0.602814988579542, + "grad_norm": 0.4069460332393646, + "learning_rate": 8.064753013187522e-05, + "loss": 0.0521, + "step": 19530 + }, + { + "epoch": 0.6031236496080005, + "grad_norm": 0.3306931257247925, + "learning_rate": 8.062575145124289e-05, + "loss": 0.0451, + "step": 19540 + }, + { + "epoch": 0.6034323106364591, + "grad_norm": 0.34667617082595825, + "learning_rate": 8.060396346731587e-05, + "loss": 0.052, + "step": 19550 + }, + { + "epoch": 0.6037409716649176, + "grad_norm": 0.2884424924850464, + "learning_rate": 8.058216618671281e-05, + "loss": 0.0489, + "step": 19560 + }, + { + "epoch": 0.6040496326933761, + "grad_norm": 0.3618551194667816, + "learning_rate": 8.056035961605514e-05, + "loss": 0.0518, + "step": 19570 + }, + { + "epoch": 0.6043582937218347, + "grad_norm": 0.31859105825424194, + "learning_rate": 8.05385437619671e-05, + "loss": 0.0502, + "step": 19580 + }, + { + "epoch": 0.6046669547502932, + "grad_norm": 0.3457159996032715, + "learning_rate": 8.05167186310758e-05, + "loss": 0.047, + "step": 19590 + }, + { + "epoch": 0.6049756157787518, + "grad_norm": 0.36075714230537415, + "learning_rate": 8.049488423001113e-05, + "loss": 0.0494, + "step": 19600 + }, + { + "epoch": 0.6052842768072103, + "grad_norm": 0.5449519753456116, + "learning_rate": 8.047304056540581e-05, + "loss": 0.0531, + "step": 19610 + }, + { + "epoch": 0.6055929378356689, + "grad_norm": 0.3698030114173889, + "learning_rate": 8.045118764389534e-05, + "loss": 0.0502, + "step": 19620 + }, + { + "epoch": 0.6059015988641274, + "grad_norm": 0.5167145729064941, + "learning_rate": 8.042932547211809e-05, + "loss": 0.0509, + "step": 19630 + }, + { + "epoch": 0.6062102598925859, + "grad_norm": 0.29102542996406555, + "learning_rate": 8.04074540567152e-05, + "loss": 0.0486, + "step": 19640 + }, + { + "epoch": 0.6065189209210445, + "grad_norm": 0.318135142326355, + "learning_rate": 8.038557340433063e-05, + "loss": 0.0485, + "step": 19650 + }, + { + "epoch": 0.606827581949503, + "grad_norm": 0.3857664167881012, + "learning_rate": 8.036368352161115e-05, + "loss": 0.0537, + "step": 19660 + }, + { + "epoch": 0.6071362429779616, + "grad_norm": 0.6691070795059204, + "learning_rate": 8.034178441520633e-05, + "loss": 0.0534, + "step": 19670 + }, + { + "epoch": 0.6074449040064201, + "grad_norm": 0.36909881234169006, + "learning_rate": 8.031987609176852e-05, + "loss": 0.0484, + "step": 19680 + }, + { + "epoch": 0.6077535650348787, + "grad_norm": 0.2990051507949829, + "learning_rate": 8.02979585579529e-05, + "loss": 0.0461, + "step": 19690 + }, + { + "epoch": 0.6080622260633373, + "grad_norm": 0.33961278200149536, + "learning_rate": 8.027603182041745e-05, + "loss": 0.0472, + "step": 19700 + }, + { + "epoch": 0.6083708870917958, + "grad_norm": 0.3369821310043335, + "learning_rate": 8.025409588582292e-05, + "loss": 0.0473, + "step": 19710 + }, + { + "epoch": 0.6086795481202544, + "grad_norm": 0.39961573481559753, + "learning_rate": 8.023215076083288e-05, + "loss": 0.052, + "step": 19720 + }, + { + "epoch": 0.6089882091487129, + "grad_norm": 0.35704153776168823, + "learning_rate": 8.021019645211367e-05, + "loss": 0.05, + "step": 19730 + }, + { + "epoch": 0.6092968701771714, + "grad_norm": 0.283109188079834, + "learning_rate": 8.018823296633441e-05, + "loss": 0.0525, + "step": 19740 + }, + { + "epoch": 0.60960553120563, + "grad_norm": 0.46085989475250244, + "learning_rate": 8.016626031016708e-05, + "loss": 0.0502, + "step": 19750 + }, + { + "epoch": 0.6099141922340885, + "grad_norm": 0.4136348068714142, + "learning_rate": 8.014427849028636e-05, + "loss": 0.0524, + "step": 19760 + }, + { + "epoch": 0.6102228532625471, + "grad_norm": 0.3569639027118683, + "learning_rate": 8.012228751336974e-05, + "loss": 0.0507, + "step": 19770 + }, + { + "epoch": 0.6105315142910056, + "grad_norm": 0.35549530386924744, + "learning_rate": 8.01002873860975e-05, + "loss": 0.0508, + "step": 19780 + }, + { + "epoch": 0.6108401753194642, + "grad_norm": 0.38731858134269714, + "learning_rate": 8.00782781151527e-05, + "loss": 0.0533, + "step": 19790 + }, + { + "epoch": 0.6111488363479227, + "grad_norm": 0.3112386167049408, + "learning_rate": 8.005625970722119e-05, + "loss": 0.0526, + "step": 19800 + }, + { + "epoch": 0.6114574973763812, + "grad_norm": 0.485972136259079, + "learning_rate": 8.003423216899158e-05, + "loss": 0.0479, + "step": 19810 + }, + { + "epoch": 0.6117661584048398, + "grad_norm": 0.279873788356781, + "learning_rate": 8.001219550715522e-05, + "loss": 0.0502, + "step": 19820 + }, + { + "epoch": 0.6120748194332983, + "grad_norm": 0.2958970367908478, + "learning_rate": 7.999014972840632e-05, + "loss": 0.0518, + "step": 19830 + }, + { + "epoch": 0.612383480461757, + "grad_norm": 0.3222665786743164, + "learning_rate": 7.996809483944174e-05, + "loss": 0.0472, + "step": 19840 + }, + { + "epoch": 0.6126921414902154, + "grad_norm": 0.3042316138744354, + "learning_rate": 7.994603084696124e-05, + "loss": 0.0521, + "step": 19850 + }, + { + "epoch": 0.613000802518674, + "grad_norm": 0.3169701099395752, + "learning_rate": 7.992395775766724e-05, + "loss": 0.0492, + "step": 19860 + }, + { + "epoch": 0.6133094635471326, + "grad_norm": 0.3243609070777893, + "learning_rate": 7.990187557826497e-05, + "loss": 0.0511, + "step": 19870 + }, + { + "epoch": 0.6136181245755911, + "grad_norm": 0.3365952670574188, + "learning_rate": 7.987978431546242e-05, + "loss": 0.0489, + "step": 19880 + }, + { + "epoch": 0.6139267856040497, + "grad_norm": 0.26862651109695435, + "learning_rate": 7.985768397597031e-05, + "loss": 0.049, + "step": 19890 + }, + { + "epoch": 0.6142354466325082, + "grad_norm": 0.37664371728897095, + "learning_rate": 7.983557456650216e-05, + "loss": 0.0475, + "step": 19900 + }, + { + "epoch": 0.6145441076609667, + "grad_norm": 0.42740291357040405, + "learning_rate": 7.981345609377422e-05, + "loss": 0.0515, + "step": 19910 + }, + { + "epoch": 0.6148527686894253, + "grad_norm": 0.38409098982810974, + "learning_rate": 7.97913285645055e-05, + "loss": 0.0465, + "step": 19920 + }, + { + "epoch": 0.6151614297178838, + "grad_norm": 0.36181989312171936, + "learning_rate": 7.976919198541776e-05, + "loss": 0.0565, + "step": 19930 + }, + { + "epoch": 0.6154700907463424, + "grad_norm": 0.33922237157821655, + "learning_rate": 7.974704636323548e-05, + "loss": 0.0484, + "step": 19940 + }, + { + "epoch": 0.6157787517748009, + "grad_norm": 0.45795080065727234, + "learning_rate": 7.972489170468597e-05, + "loss": 0.0493, + "step": 19950 + }, + { + "epoch": 0.6160874128032595, + "grad_norm": 0.4939574599266052, + "learning_rate": 7.970272801649918e-05, + "loss": 0.055, + "step": 19960 + }, + { + "epoch": 0.616396073831718, + "grad_norm": 0.3242253363132477, + "learning_rate": 7.96805553054079e-05, + "loss": 0.0478, + "step": 19970 + }, + { + "epoch": 0.6167047348601765, + "grad_norm": 0.4346364438533783, + "learning_rate": 7.965837357814756e-05, + "loss": 0.051, + "step": 19980 + }, + { + "epoch": 0.6170133958886351, + "grad_norm": 0.45463886857032776, + "learning_rate": 7.963618284145643e-05, + "loss": 0.0458, + "step": 19990 + }, + { + "epoch": 0.6173220569170936, + "grad_norm": 0.3130718171596527, + "learning_rate": 7.961398310207544e-05, + "loss": 0.0508, + "step": 20000 + }, + { + "epoch": 0.6176307179455522, + "grad_norm": 0.3019697666168213, + "learning_rate": 7.95917743667483e-05, + "loss": 0.0486, + "step": 20010 + }, + { + "epoch": 0.6179393789740107, + "grad_norm": 0.3967873752117157, + "learning_rate": 7.956955664222144e-05, + "loss": 0.0494, + "step": 20020 + }, + { + "epoch": 0.6182480400024692, + "grad_norm": 0.3511422276496887, + "learning_rate": 7.954732993524399e-05, + "loss": 0.0527, + "step": 20030 + }, + { + "epoch": 0.6185567010309279, + "grad_norm": 0.48957616090774536, + "learning_rate": 7.952509425256786e-05, + "loss": 0.0485, + "step": 20040 + }, + { + "epoch": 0.6188653620593864, + "grad_norm": 0.24779662489891052, + "learning_rate": 7.950284960094767e-05, + "loss": 0.046, + "step": 20050 + }, + { + "epoch": 0.619174023087845, + "grad_norm": 0.2531059980392456, + "learning_rate": 7.948059598714076e-05, + "loss": 0.0461, + "step": 20060 + }, + { + "epoch": 0.6194826841163035, + "grad_norm": 0.5524089336395264, + "learning_rate": 7.945833341790717e-05, + "loss": 0.0472, + "step": 20070 + }, + { + "epoch": 0.619791345144762, + "grad_norm": 0.4370622932910919, + "learning_rate": 7.94360619000097e-05, + "loss": 0.0475, + "step": 20080 + }, + { + "epoch": 0.6201000061732206, + "grad_norm": 0.28986647725105286, + "learning_rate": 7.941378144021381e-05, + "loss": 0.0497, + "step": 20090 + }, + { + "epoch": 0.6204086672016791, + "grad_norm": 0.29568588733673096, + "learning_rate": 7.939149204528777e-05, + "loss": 0.0463, + "step": 20100 + }, + { + "epoch": 0.6207173282301377, + "grad_norm": 0.35503828525543213, + "learning_rate": 7.936919372200246e-05, + "loss": 0.0502, + "step": 20110 + }, + { + "epoch": 0.6210259892585962, + "grad_norm": 0.33737313747406006, + "learning_rate": 7.934688647713158e-05, + "loss": 0.0487, + "step": 20120 + }, + { + "epoch": 0.6213346502870547, + "grad_norm": 0.3389264941215515, + "learning_rate": 7.932457031745143e-05, + "loss": 0.0496, + "step": 20130 + }, + { + "epoch": 0.6216433113155133, + "grad_norm": 0.3242708742618561, + "learning_rate": 7.930224524974108e-05, + "loss": 0.0517, + "step": 20140 + }, + { + "epoch": 0.6219519723439718, + "grad_norm": 0.40020760893821716, + "learning_rate": 7.927991128078232e-05, + "loss": 0.0464, + "step": 20150 + }, + { + "epoch": 0.6222606333724304, + "grad_norm": 0.3466799259185791, + "learning_rate": 7.925756841735958e-05, + "loss": 0.0545, + "step": 20160 + }, + { + "epoch": 0.6225692944008889, + "grad_norm": 0.43135321140289307, + "learning_rate": 7.923521666626008e-05, + "loss": 0.0491, + "step": 20170 + }, + { + "epoch": 0.6228779554293475, + "grad_norm": 0.6236898303031921, + "learning_rate": 7.921285603427366e-05, + "loss": 0.0489, + "step": 20180 + }, + { + "epoch": 0.623186616457806, + "grad_norm": 0.2641715407371521, + "learning_rate": 7.91904865281929e-05, + "loss": 0.0516, + "step": 20190 + }, + { + "epoch": 0.6234952774862645, + "grad_norm": 0.33305999636650085, + "learning_rate": 7.916810815481307e-05, + "loss": 0.0503, + "step": 20200 + }, + { + "epoch": 0.6238039385147232, + "grad_norm": 0.33465009927749634, + "learning_rate": 7.914572092093211e-05, + "loss": 0.0567, + "step": 20210 + }, + { + "epoch": 0.6241125995431817, + "grad_norm": 0.4132522642612457, + "learning_rate": 7.912332483335068e-05, + "loss": 0.0542, + "step": 20220 + }, + { + "epoch": 0.6244212605716403, + "grad_norm": 0.3291358947753906, + "learning_rate": 7.910091989887213e-05, + "loss": 0.0511, + "step": 20230 + }, + { + "epoch": 0.6247299216000988, + "grad_norm": 0.22515061497688293, + "learning_rate": 7.907850612430248e-05, + "loss": 0.05, + "step": 20240 + }, + { + "epoch": 0.6250385826285573, + "grad_norm": 0.36471301317214966, + "learning_rate": 7.905608351645044e-05, + "loss": 0.0477, + "step": 20250 + }, + { + "epoch": 0.6253472436570159, + "grad_norm": 0.46685388684272766, + "learning_rate": 7.90336520821274e-05, + "loss": 0.0516, + "step": 20260 + }, + { + "epoch": 0.6256559046854744, + "grad_norm": 0.3629143238067627, + "learning_rate": 7.901121182814746e-05, + "loss": 0.0492, + "step": 20270 + }, + { + "epoch": 0.625964565713933, + "grad_norm": 0.4246806204319, + "learning_rate": 7.898876276132736e-05, + "loss": 0.0514, + "step": 20280 + }, + { + "epoch": 0.6262732267423915, + "grad_norm": 0.39314568042755127, + "learning_rate": 7.896630488848654e-05, + "loss": 0.0502, + "step": 20290 + }, + { + "epoch": 0.62658188777085, + "grad_norm": 0.35658779740333557, + "learning_rate": 7.89438382164471e-05, + "loss": 0.05, + "step": 20300 + }, + { + "epoch": 0.6268905487993086, + "grad_norm": 0.33215606212615967, + "learning_rate": 7.892136275203383e-05, + "loss": 0.0513, + "step": 20310 + }, + { + "epoch": 0.6271992098277671, + "grad_norm": 0.2715575695037842, + "learning_rate": 7.889887850207418e-05, + "loss": 0.0453, + "step": 20320 + }, + { + "epoch": 0.6275078708562257, + "grad_norm": 0.31782227754592896, + "learning_rate": 7.887638547339827e-05, + "loss": 0.0506, + "step": 20330 + }, + { + "epoch": 0.6278165318846842, + "grad_norm": 0.31076011061668396, + "learning_rate": 7.885388367283891e-05, + "loss": 0.0462, + "step": 20340 + }, + { + "epoch": 0.6281251929131428, + "grad_norm": 0.4148462414741516, + "learning_rate": 7.88313731072315e-05, + "loss": 0.0474, + "step": 20350 + }, + { + "epoch": 0.6284338539416013, + "grad_norm": 0.2828272879123688, + "learning_rate": 7.88088537834142e-05, + "loss": 0.0539, + "step": 20360 + }, + { + "epoch": 0.6287425149700598, + "grad_norm": 0.26057612895965576, + "learning_rate": 7.878632570822778e-05, + "loss": 0.0469, + "step": 20370 + }, + { + "epoch": 0.6290511759985185, + "grad_norm": 0.6095553636550903, + "learning_rate": 7.876378888851567e-05, + "loss": 0.0537, + "step": 20380 + }, + { + "epoch": 0.629359837026977, + "grad_norm": 0.7107054591178894, + "learning_rate": 7.874124333112396e-05, + "loss": 0.0468, + "step": 20390 + }, + { + "epoch": 0.6296684980554356, + "grad_norm": 0.4027869701385498, + "learning_rate": 7.871868904290138e-05, + "loss": 0.0526, + "step": 20400 + }, + { + "epoch": 0.6299771590838941, + "grad_norm": 0.5037659406661987, + "learning_rate": 7.869612603069935e-05, + "loss": 0.0479, + "step": 20410 + }, + { + "epoch": 0.6302858201123526, + "grad_norm": 0.34722623229026794, + "learning_rate": 7.867355430137192e-05, + "loss": 0.0466, + "step": 20420 + }, + { + "epoch": 0.6305944811408112, + "grad_norm": 0.2996414601802826, + "learning_rate": 7.865097386177577e-05, + "loss": 0.0468, + "step": 20430 + }, + { + "epoch": 0.6309031421692697, + "grad_norm": 0.9735236763954163, + "learning_rate": 7.862838471877023e-05, + "loss": 0.0513, + "step": 20440 + }, + { + "epoch": 0.6312118031977283, + "grad_norm": 0.540736198425293, + "learning_rate": 7.860578687921731e-05, + "loss": 0.0492, + "step": 20450 + }, + { + "epoch": 0.6315204642261868, + "grad_norm": 0.3022603988647461, + "learning_rate": 7.858318034998164e-05, + "loss": 0.048, + "step": 20460 + }, + { + "epoch": 0.6318291252546453, + "grad_norm": 0.3984932005405426, + "learning_rate": 7.856056513793046e-05, + "loss": 0.0483, + "step": 20470 + }, + { + "epoch": 0.6321377862831039, + "grad_norm": 0.259240061044693, + "learning_rate": 7.85379412499337e-05, + "loss": 0.0522, + "step": 20480 + }, + { + "epoch": 0.6324464473115624, + "grad_norm": 0.25300082564353943, + "learning_rate": 7.851530869286389e-05, + "loss": 0.0511, + "step": 20490 + }, + { + "epoch": 0.632755108340021, + "grad_norm": 0.3391929268836975, + "learning_rate": 7.849266747359619e-05, + "loss": 0.0437, + "step": 20500 + }, + { + "epoch": 0.6330637693684795, + "grad_norm": 0.3435535430908203, + "learning_rate": 7.847001759900843e-05, + "loss": 0.05, + "step": 20510 + }, + { + "epoch": 0.6333724303969381, + "grad_norm": 0.3685261607170105, + "learning_rate": 7.844735907598102e-05, + "loss": 0.045, + "step": 20520 + }, + { + "epoch": 0.6336810914253966, + "grad_norm": 0.3062552511692047, + "learning_rate": 7.842469191139703e-05, + "loss": 0.0517, + "step": 20530 + }, + { + "epoch": 0.6339897524538551, + "grad_norm": 0.39745330810546875, + "learning_rate": 7.840201611214215e-05, + "loss": 0.0518, + "step": 20540 + }, + { + "epoch": 0.6342984134823137, + "grad_norm": 0.3241347074508667, + "learning_rate": 7.837933168510469e-05, + "loss": 0.0459, + "step": 20550 + }, + { + "epoch": 0.6346070745107723, + "grad_norm": 0.38086944818496704, + "learning_rate": 7.835663863717559e-05, + "loss": 0.0473, + "step": 20560 + }, + { + "epoch": 0.6349157355392309, + "grad_norm": 0.29067134857177734, + "learning_rate": 7.833393697524838e-05, + "loss": 0.0471, + "step": 20570 + }, + { + "epoch": 0.6352243965676894, + "grad_norm": 0.39038291573524475, + "learning_rate": 7.831122670621922e-05, + "loss": 0.0516, + "step": 20580 + }, + { + "epoch": 0.6355330575961479, + "grad_norm": 0.3918664753437042, + "learning_rate": 7.82885078369869e-05, + "loss": 0.0501, + "step": 20590 + }, + { + "epoch": 0.6358417186246065, + "grad_norm": 0.3207690119743347, + "learning_rate": 7.826578037445283e-05, + "loss": 0.0485, + "step": 20600 + }, + { + "epoch": 0.636150379653065, + "grad_norm": 0.35032618045806885, + "learning_rate": 7.824304432552097e-05, + "loss": 0.0484, + "step": 20610 + }, + { + "epoch": 0.6364590406815236, + "grad_norm": 0.30847224593162537, + "learning_rate": 7.822029969709798e-05, + "loss": 0.0453, + "step": 20620 + }, + { + "epoch": 0.6367677017099821, + "grad_norm": 0.2813892960548401, + "learning_rate": 7.819754649609306e-05, + "loss": 0.0474, + "step": 20630 + }, + { + "epoch": 0.6370763627384406, + "grad_norm": 0.3619902729988098, + "learning_rate": 7.817478472941802e-05, + "loss": 0.0497, + "step": 20640 + }, + { + "epoch": 0.6373850237668992, + "grad_norm": 0.4993097186088562, + "learning_rate": 7.815201440398727e-05, + "loss": 0.047, + "step": 20650 + }, + { + "epoch": 0.6376936847953577, + "grad_norm": 0.39735421538352966, + "learning_rate": 7.812923552671789e-05, + "loss": 0.05, + "step": 20660 + }, + { + "epoch": 0.6380023458238163, + "grad_norm": 0.3734968602657318, + "learning_rate": 7.810644810452945e-05, + "loss": 0.0461, + "step": 20670 + }, + { + "epoch": 0.6383110068522748, + "grad_norm": 0.376437783241272, + "learning_rate": 7.808365214434417e-05, + "loss": 0.0502, + "step": 20680 + }, + { + "epoch": 0.6386196678807334, + "grad_norm": 0.3059508204460144, + "learning_rate": 7.80608476530869e-05, + "loss": 0.0533, + "step": 20690 + }, + { + "epoch": 0.6389283289091919, + "grad_norm": 0.31573405861854553, + "learning_rate": 7.8038034637685e-05, + "loss": 0.0489, + "step": 20700 + }, + { + "epoch": 0.6392369899376504, + "grad_norm": 0.31935110688209534, + "learning_rate": 7.801521310506848e-05, + "loss": 0.0506, + "step": 20710 + }, + { + "epoch": 0.639545650966109, + "grad_norm": 0.26701217889785767, + "learning_rate": 7.799238306216994e-05, + "loss": 0.052, + "step": 20720 + }, + { + "epoch": 0.6398543119945675, + "grad_norm": 0.40284740924835205, + "learning_rate": 7.796954451592448e-05, + "loss": 0.0519, + "step": 20730 + }, + { + "epoch": 0.6401629730230262, + "grad_norm": 0.4114428758621216, + "learning_rate": 7.794669747326992e-05, + "loss": 0.0445, + "step": 20740 + }, + { + "epoch": 0.6404716340514847, + "grad_norm": 0.3534519374370575, + "learning_rate": 7.792384194114654e-05, + "loss": 0.0483, + "step": 20750 + }, + { + "epoch": 0.6407802950799432, + "grad_norm": 0.3267820179462433, + "learning_rate": 7.790097792649729e-05, + "loss": 0.0449, + "step": 20760 + }, + { + "epoch": 0.6410889561084018, + "grad_norm": 0.7206806540489197, + "learning_rate": 7.787810543626762e-05, + "loss": 0.0506, + "step": 20770 + }, + { + "epoch": 0.6413976171368603, + "grad_norm": 0.2737559676170349, + "learning_rate": 7.785522447740558e-05, + "loss": 0.0474, + "step": 20780 + }, + { + "epoch": 0.6417062781653189, + "grad_norm": 0.6046457290649414, + "learning_rate": 7.783233505686182e-05, + "loss": 0.0469, + "step": 20790 + }, + { + "epoch": 0.6420149391937774, + "grad_norm": 0.3015018105506897, + "learning_rate": 7.780943718158955e-05, + "loss": 0.0465, + "step": 20800 + }, + { + "epoch": 0.6423236002222359, + "grad_norm": 0.4941875636577606, + "learning_rate": 7.778653085854453e-05, + "loss": 0.0489, + "step": 20810 + }, + { + "epoch": 0.6426322612506945, + "grad_norm": 0.28531813621520996, + "learning_rate": 7.77636160946851e-05, + "loss": 0.0489, + "step": 20820 + }, + { + "epoch": 0.642940922279153, + "grad_norm": 0.3109099566936493, + "learning_rate": 7.774069289697215e-05, + "loss": 0.05, + "step": 20830 + }, + { + "epoch": 0.6432495833076116, + "grad_norm": 0.330633282661438, + "learning_rate": 7.771776127236913e-05, + "loss": 0.0433, + "step": 20840 + }, + { + "epoch": 0.6435582443360701, + "grad_norm": 0.8814764022827148, + "learning_rate": 7.769482122784212e-05, + "loss": 0.048, + "step": 20850 + }, + { + "epoch": 0.6438669053645287, + "grad_norm": 0.3210153579711914, + "learning_rate": 7.767187277035963e-05, + "loss": 0.0512, + "step": 20860 + }, + { + "epoch": 0.6441755663929872, + "grad_norm": 0.406869500875473, + "learning_rate": 7.764891590689285e-05, + "loss": 0.0538, + "step": 20870 + }, + { + "epoch": 0.6444842274214457, + "grad_norm": 0.3948357403278351, + "learning_rate": 7.762595064441542e-05, + "loss": 0.0472, + "step": 20880 + }, + { + "epoch": 0.6447928884499043, + "grad_norm": 0.3061201572418213, + "learning_rate": 7.760297698990362e-05, + "loss": 0.0504, + "step": 20890 + }, + { + "epoch": 0.6451015494783628, + "grad_norm": 0.45533618330955505, + "learning_rate": 7.757999495033623e-05, + "loss": 0.0529, + "step": 20900 + }, + { + "epoch": 0.6454102105068215, + "grad_norm": 0.287897527217865, + "learning_rate": 7.755700453269456e-05, + "loss": 0.0487, + "step": 20910 + }, + { + "epoch": 0.64571887153528, + "grad_norm": 0.3917054831981659, + "learning_rate": 7.753400574396254e-05, + "loss": 0.0518, + "step": 20920 + }, + { + "epoch": 0.6460275325637385, + "grad_norm": 0.2869715094566345, + "learning_rate": 7.751099859112655e-05, + "loss": 0.0468, + "step": 20930 + }, + { + "epoch": 0.6463361935921971, + "grad_norm": 0.4361961781978607, + "learning_rate": 7.748798308117557e-05, + "loss": 0.0493, + "step": 20940 + }, + { + "epoch": 0.6466448546206556, + "grad_norm": 0.3560357987880707, + "learning_rate": 7.746495922110112e-05, + "loss": 0.0527, + "step": 20950 + }, + { + "epoch": 0.6469535156491142, + "grad_norm": 0.4387330710887909, + "learning_rate": 7.744192701789723e-05, + "loss": 0.0534, + "step": 20960 + }, + { + "epoch": 0.6472621766775727, + "grad_norm": 0.3576853573322296, + "learning_rate": 7.741888647856046e-05, + "loss": 0.0499, + "step": 20970 + }, + { + "epoch": 0.6475708377060312, + "grad_norm": 0.27785542607307434, + "learning_rate": 7.739583761008994e-05, + "loss": 0.0518, + "step": 20980 + }, + { + "epoch": 0.6478794987344898, + "grad_norm": 0.4122113883495331, + "learning_rate": 7.73727804194873e-05, + "loss": 0.05, + "step": 20990 + }, + { + "epoch": 0.6481881597629483, + "grad_norm": 0.26664403080940247, + "learning_rate": 7.734971491375671e-05, + "loss": 0.0454, + "step": 21000 + }, + { + "epoch": 0.6484968207914069, + "grad_norm": 0.37241336703300476, + "learning_rate": 7.732664109990485e-05, + "loss": 0.0536, + "step": 21010 + }, + { + "epoch": 0.6488054818198654, + "grad_norm": 0.3249344527721405, + "learning_rate": 7.730355898494095e-05, + "loss": 0.0511, + "step": 21020 + }, + { + "epoch": 0.6491141428483239, + "grad_norm": 0.3596643805503845, + "learning_rate": 7.728046857587673e-05, + "loss": 0.0493, + "step": 21030 + }, + { + "epoch": 0.6494228038767825, + "grad_norm": 0.32462090253829956, + "learning_rate": 7.725736987972647e-05, + "loss": 0.0505, + "step": 21040 + }, + { + "epoch": 0.649731464905241, + "grad_norm": 0.4498898983001709, + "learning_rate": 7.723426290350691e-05, + "loss": 0.0513, + "step": 21050 + }, + { + "epoch": 0.6500401259336996, + "grad_norm": 0.4134032428264618, + "learning_rate": 7.721114765423736e-05, + "loss": 0.0503, + "step": 21060 + }, + { + "epoch": 0.6503487869621581, + "grad_norm": 0.32947495579719543, + "learning_rate": 7.718802413893963e-05, + "loss": 0.0429, + "step": 21070 + }, + { + "epoch": 0.6506574479906168, + "grad_norm": 0.36131751537323, + "learning_rate": 7.716489236463802e-05, + "loss": 0.0486, + "step": 21080 + }, + { + "epoch": 0.6509661090190753, + "grad_norm": 0.3147152066230774, + "learning_rate": 7.714175233835936e-05, + "loss": 0.049, + "step": 21090 + }, + { + "epoch": 0.6512747700475338, + "grad_norm": 0.3118889033794403, + "learning_rate": 7.711860406713299e-05, + "loss": 0.0497, + "step": 21100 + }, + { + "epoch": 0.6515834310759924, + "grad_norm": 0.4690038561820984, + "learning_rate": 7.70954475579907e-05, + "loss": 0.0476, + "step": 21110 + }, + { + "epoch": 0.6518920921044509, + "grad_norm": 0.32995009422302246, + "learning_rate": 7.707228281796688e-05, + "loss": 0.0494, + "step": 21120 + }, + { + "epoch": 0.6522007531329095, + "grad_norm": 0.5725066065788269, + "learning_rate": 7.704910985409833e-05, + "loss": 0.0552, + "step": 21130 + }, + { + "epoch": 0.652509414161368, + "grad_norm": 0.6677294969558716, + "learning_rate": 7.702592867342439e-05, + "loss": 0.0471, + "step": 21140 + }, + { + "epoch": 0.6528180751898265, + "grad_norm": 0.31844520568847656, + "learning_rate": 7.700273928298691e-05, + "loss": 0.0502, + "step": 21150 + }, + { + "epoch": 0.6531267362182851, + "grad_norm": 0.397108793258667, + "learning_rate": 7.697954168983021e-05, + "loss": 0.047, + "step": 21160 + }, + { + "epoch": 0.6534353972467436, + "grad_norm": 0.3335418701171875, + "learning_rate": 7.695633590100109e-05, + "loss": 0.0507, + "step": 21170 + }, + { + "epoch": 0.6537440582752022, + "grad_norm": 0.32124030590057373, + "learning_rate": 7.693312192354886e-05, + "loss": 0.0485, + "step": 21180 + }, + { + "epoch": 0.6540527193036607, + "grad_norm": 0.3745575249195099, + "learning_rate": 7.690989976452532e-05, + "loss": 0.0518, + "step": 21190 + }, + { + "epoch": 0.6543613803321192, + "grad_norm": 0.30660200119018555, + "learning_rate": 7.688666943098475e-05, + "loss": 0.0459, + "step": 21200 + }, + { + "epoch": 0.6546700413605778, + "grad_norm": 0.366639643907547, + "learning_rate": 7.686343092998389e-05, + "loss": 0.0483, + "step": 21210 + }, + { + "epoch": 0.6549787023890363, + "grad_norm": 0.6222352385520935, + "learning_rate": 7.684018426858202e-05, + "loss": 0.0463, + "step": 21220 + }, + { + "epoch": 0.6552873634174949, + "grad_norm": 0.35977381467819214, + "learning_rate": 7.681692945384084e-05, + "loss": 0.0517, + "step": 21230 + }, + { + "epoch": 0.6555960244459534, + "grad_norm": 0.374286025762558, + "learning_rate": 7.679366649282456e-05, + "loss": 0.0511, + "step": 21240 + }, + { + "epoch": 0.655904685474412, + "grad_norm": 0.2696363031864166, + "learning_rate": 7.677039539259983e-05, + "loss": 0.046, + "step": 21250 + }, + { + "epoch": 0.6562133465028706, + "grad_norm": 0.3433956801891327, + "learning_rate": 7.674711616023581e-05, + "loss": 0.0446, + "step": 21260 + }, + { + "epoch": 0.656522007531329, + "grad_norm": 0.4746894836425781, + "learning_rate": 7.672382880280413e-05, + "loss": 0.0487, + "step": 21270 + }, + { + "epoch": 0.6568306685597877, + "grad_norm": 0.3003491461277008, + "learning_rate": 7.670053332737885e-05, + "loss": 0.0458, + "step": 21280 + }, + { + "epoch": 0.6571393295882462, + "grad_norm": 0.42890849709510803, + "learning_rate": 7.667722974103654e-05, + "loss": 0.0517, + "step": 21290 + }, + { + "epoch": 0.6574479906167048, + "grad_norm": 0.31265342235565186, + "learning_rate": 7.66539180508562e-05, + "loss": 0.0489, + "step": 21300 + }, + { + "epoch": 0.6577566516451633, + "grad_norm": 0.41905900835990906, + "learning_rate": 7.663059826391932e-05, + "loss": 0.0544, + "step": 21310 + }, + { + "epoch": 0.6580653126736218, + "grad_norm": 0.8424597978591919, + "learning_rate": 7.660727038730981e-05, + "loss": 0.0465, + "step": 21320 + }, + { + "epoch": 0.6583739737020804, + "grad_norm": 0.447329044342041, + "learning_rate": 7.65839344281141e-05, + "loss": 0.0479, + "step": 21330 + }, + { + "epoch": 0.6586826347305389, + "grad_norm": 0.3497753441333771, + "learning_rate": 7.656059039342101e-05, + "loss": 0.0531, + "step": 21340 + }, + { + "epoch": 0.6589912957589975, + "grad_norm": 0.281967431306839, + "learning_rate": 7.653723829032187e-05, + "loss": 0.0505, + "step": 21350 + }, + { + "epoch": 0.659299956787456, + "grad_norm": 0.28999781608581543, + "learning_rate": 7.65138781259104e-05, + "loss": 0.0506, + "step": 21360 + }, + { + "epoch": 0.6596086178159145, + "grad_norm": 0.29247379302978516, + "learning_rate": 7.649050990728279e-05, + "loss": 0.0443, + "step": 21370 + }, + { + "epoch": 0.6599172788443731, + "grad_norm": 0.39541491866111755, + "learning_rate": 7.646713364153774e-05, + "loss": 0.048, + "step": 21380 + }, + { + "epoch": 0.6602259398728316, + "grad_norm": 0.4130237102508545, + "learning_rate": 7.64437493357763e-05, + "loss": 0.046, + "step": 21390 + }, + { + "epoch": 0.6605346009012902, + "grad_norm": 0.3432020843029022, + "learning_rate": 7.642035699710202e-05, + "loss": 0.049, + "step": 21400 + }, + { + "epoch": 0.6608432619297487, + "grad_norm": 0.32062700390815735, + "learning_rate": 7.639695663262089e-05, + "loss": 0.0491, + "step": 21410 + }, + { + "epoch": 0.6611519229582073, + "grad_norm": 0.4147270619869232, + "learning_rate": 7.637354824944128e-05, + "loss": 0.049, + "step": 21420 + }, + { + "epoch": 0.6614605839866659, + "grad_norm": 0.3636077046394348, + "learning_rate": 7.635013185467408e-05, + "loss": 0.0454, + "step": 21430 + }, + { + "epoch": 0.6617692450151244, + "grad_norm": 0.8679901361465454, + "learning_rate": 7.632670745543256e-05, + "loss": 0.0472, + "step": 21440 + }, + { + "epoch": 0.662077906043583, + "grad_norm": 0.413166344165802, + "learning_rate": 7.630327505883242e-05, + "loss": 0.0482, + "step": 21450 + }, + { + "epoch": 0.6623865670720415, + "grad_norm": 0.5529443025588989, + "learning_rate": 7.627983467199182e-05, + "loss": 0.0486, + "step": 21460 + }, + { + "epoch": 0.6626952281005001, + "grad_norm": 0.31937071681022644, + "learning_rate": 7.625638630203132e-05, + "loss": 0.0507, + "step": 21470 + }, + { + "epoch": 0.6630038891289586, + "grad_norm": 0.37978705763816833, + "learning_rate": 7.623292995607394e-05, + "loss": 0.0469, + "step": 21480 + }, + { + "epoch": 0.6633125501574171, + "grad_norm": 0.3322894275188446, + "learning_rate": 7.620946564124507e-05, + "loss": 0.0511, + "step": 21490 + }, + { + "epoch": 0.6636212111858757, + "grad_norm": 0.24674317240715027, + "learning_rate": 7.618599336467256e-05, + "loss": 0.0445, + "step": 21500 + }, + { + "epoch": 0.6639298722143342, + "grad_norm": 0.3003174364566803, + "learning_rate": 7.616251313348666e-05, + "loss": 0.0507, + "step": 21510 + }, + { + "epoch": 0.6642385332427928, + "grad_norm": 0.4011954665184021, + "learning_rate": 7.613902495482005e-05, + "loss": 0.0554, + "step": 21520 + }, + { + "epoch": 0.6645471942712513, + "grad_norm": 0.6259546875953674, + "learning_rate": 7.611552883580784e-05, + "loss": 0.0469, + "step": 21530 + }, + { + "epoch": 0.6648558552997098, + "grad_norm": 0.3225897252559662, + "learning_rate": 7.609202478358748e-05, + "loss": 0.0436, + "step": 21540 + }, + { + "epoch": 0.6651645163281684, + "grad_norm": 0.41534993052482605, + "learning_rate": 7.606851280529895e-05, + "loss": 0.0492, + "step": 21550 + }, + { + "epoch": 0.6654731773566269, + "grad_norm": 0.4353807270526886, + "learning_rate": 7.604499290808449e-05, + "loss": 0.0489, + "step": 21560 + }, + { + "epoch": 0.6657818383850855, + "grad_norm": 0.24049992859363556, + "learning_rate": 7.602146509908888e-05, + "loss": 0.0486, + "step": 21570 + }, + { + "epoch": 0.666090499413544, + "grad_norm": 0.3256949186325073, + "learning_rate": 7.599792938545921e-05, + "loss": 0.0481, + "step": 21580 + }, + { + "epoch": 0.6663991604420026, + "grad_norm": 0.3669275641441345, + "learning_rate": 7.597438577434506e-05, + "loss": 0.0485, + "step": 21590 + }, + { + "epoch": 0.6667078214704611, + "grad_norm": 0.36440759897232056, + "learning_rate": 7.595083427289831e-05, + "loss": 0.0438, + "step": 21600 + }, + { + "epoch": 0.6670164824989196, + "grad_norm": 0.3958587944507599, + "learning_rate": 7.59272748882733e-05, + "loss": 0.0526, + "step": 21610 + }, + { + "epoch": 0.6673251435273783, + "grad_norm": 0.37887415289878845, + "learning_rate": 7.590370762762675e-05, + "loss": 0.0471, + "step": 21620 + }, + { + "epoch": 0.6676338045558368, + "grad_norm": 0.2881532311439514, + "learning_rate": 7.588013249811777e-05, + "loss": 0.0471, + "step": 21630 + }, + { + "epoch": 0.6679424655842954, + "grad_norm": 0.3907415270805359, + "learning_rate": 7.585654950690786e-05, + "loss": 0.047, + "step": 21640 + }, + { + "epoch": 0.6682511266127539, + "grad_norm": 0.39244022965431213, + "learning_rate": 7.583295866116091e-05, + "loss": 0.0462, + "step": 21650 + }, + { + "epoch": 0.6685597876412124, + "grad_norm": 0.3581380546092987, + "learning_rate": 7.580935996804321e-05, + "loss": 0.0464, + "step": 21660 + }, + { + "epoch": 0.668868448669671, + "grad_norm": 0.3637107014656067, + "learning_rate": 7.57857534347234e-05, + "loss": 0.0481, + "step": 21670 + }, + { + "epoch": 0.6691771096981295, + "grad_norm": 0.3673078417778015, + "learning_rate": 7.576213906837254e-05, + "loss": 0.0476, + "step": 21680 + }, + { + "epoch": 0.6694857707265881, + "grad_norm": 0.342914879322052, + "learning_rate": 7.573851687616403e-05, + "loss": 0.0471, + "step": 21690 + }, + { + "epoch": 0.6697944317550466, + "grad_norm": 0.296560138463974, + "learning_rate": 7.571488686527368e-05, + "loss": 0.0501, + "step": 21700 + }, + { + "epoch": 0.6701030927835051, + "grad_norm": 0.3499191105365753, + "learning_rate": 7.569124904287968e-05, + "loss": 0.0498, + "step": 21710 + }, + { + "epoch": 0.6704117538119637, + "grad_norm": 0.4141637086868286, + "learning_rate": 7.566760341616254e-05, + "loss": 0.0482, + "step": 21720 + }, + { + "epoch": 0.6707204148404222, + "grad_norm": 0.31371164321899414, + "learning_rate": 7.564394999230519e-05, + "loss": 0.051, + "step": 21730 + }, + { + "epoch": 0.6710290758688808, + "grad_norm": 0.6780457496643066, + "learning_rate": 7.562028877849294e-05, + "loss": 0.0482, + "step": 21740 + }, + { + "epoch": 0.6713377368973393, + "grad_norm": 0.33541786670684814, + "learning_rate": 7.559661978191341e-05, + "loss": 0.0506, + "step": 21750 + }, + { + "epoch": 0.6716463979257978, + "grad_norm": 0.29629209637641907, + "learning_rate": 7.557294300975664e-05, + "loss": 0.047, + "step": 21760 + }, + { + "epoch": 0.6719550589542564, + "grad_norm": 0.44993871450424194, + "learning_rate": 7.554925846921499e-05, + "loss": 0.0486, + "step": 21770 + }, + { + "epoch": 0.672263719982715, + "grad_norm": 0.2591606676578522, + "learning_rate": 7.552556616748321e-05, + "loss": 0.0472, + "step": 21780 + }, + { + "epoch": 0.6725723810111736, + "grad_norm": 0.327762246131897, + "learning_rate": 7.550186611175838e-05, + "loss": 0.0427, + "step": 21790 + }, + { + "epoch": 0.6728810420396321, + "grad_norm": 0.29434934258461, + "learning_rate": 7.547815830923998e-05, + "loss": 0.0477, + "step": 21800 + }, + { + "epoch": 0.6731897030680907, + "grad_norm": 0.41932669281959534, + "learning_rate": 7.54544427671298e-05, + "loss": 0.0529, + "step": 21810 + }, + { + "epoch": 0.6734983640965492, + "grad_norm": 0.27874940633773804, + "learning_rate": 7.543071949263198e-05, + "loss": 0.0471, + "step": 21820 + }, + { + "epoch": 0.6738070251250077, + "grad_norm": 0.48850345611572266, + "learning_rate": 7.540698849295305e-05, + "loss": 0.0488, + "step": 21830 + }, + { + "epoch": 0.6741156861534663, + "grad_norm": 0.23310324549674988, + "learning_rate": 7.538324977530183e-05, + "loss": 0.0469, + "step": 21840 + }, + { + "epoch": 0.6744243471819248, + "grad_norm": 0.3451662063598633, + "learning_rate": 7.535950334688955e-05, + "loss": 0.0484, + "step": 21850 + }, + { + "epoch": 0.6747330082103834, + "grad_norm": 0.3999635875225067, + "learning_rate": 7.533574921492972e-05, + "loss": 0.0485, + "step": 21860 + }, + { + "epoch": 0.6750416692388419, + "grad_norm": 0.4998224079608917, + "learning_rate": 7.531198738663824e-05, + "loss": 0.0511, + "step": 21870 + }, + { + "epoch": 0.6753503302673004, + "grad_norm": 0.3384315073490143, + "learning_rate": 7.528821786923333e-05, + "loss": 0.0458, + "step": 21880 + }, + { + "epoch": 0.675658991295759, + "grad_norm": 0.42870762944221497, + "learning_rate": 7.52644406699355e-05, + "loss": 0.0492, + "step": 21890 + }, + { + "epoch": 0.6759676523242175, + "grad_norm": 0.26375725865364075, + "learning_rate": 7.524065579596766e-05, + "loss": 0.0464, + "step": 21900 + }, + { + "epoch": 0.6762763133526761, + "grad_norm": 0.2980755865573883, + "learning_rate": 7.521686325455506e-05, + "loss": 0.0467, + "step": 21910 + }, + { + "epoch": 0.6765849743811346, + "grad_norm": 0.383941113948822, + "learning_rate": 7.51930630529252e-05, + "loss": 0.0459, + "step": 21920 + }, + { + "epoch": 0.6768936354095931, + "grad_norm": 0.3637474775314331, + "learning_rate": 7.516925519830797e-05, + "loss": 0.0542, + "step": 21930 + }, + { + "epoch": 0.6772022964380517, + "grad_norm": 0.2840593457221985, + "learning_rate": 7.514543969793557e-05, + "loss": 0.0497, + "step": 21940 + }, + { + "epoch": 0.6775109574665102, + "grad_norm": 0.5057392120361328, + "learning_rate": 7.512161655904251e-05, + "loss": 0.0456, + "step": 21950 + }, + { + "epoch": 0.6778196184949689, + "grad_norm": 0.3300885856151581, + "learning_rate": 7.509778578886563e-05, + "loss": 0.0461, + "step": 21960 + }, + { + "epoch": 0.6781282795234274, + "grad_norm": 0.41024330258369446, + "learning_rate": 7.507394739464412e-05, + "loss": 0.0475, + "step": 21970 + }, + { + "epoch": 0.678436940551886, + "grad_norm": 0.25475239753723145, + "learning_rate": 7.50501013836194e-05, + "loss": 0.0473, + "step": 21980 + }, + { + "epoch": 0.6787456015803445, + "grad_norm": 0.35619843006134033, + "learning_rate": 7.50262477630353e-05, + "loss": 0.0476, + "step": 21990 + }, + { + "epoch": 0.679054262608803, + "grad_norm": 0.26911672949790955, + "learning_rate": 7.500238654013794e-05, + "loss": 0.0473, + "step": 22000 + }, + { + "epoch": 0.6793629236372616, + "grad_norm": 0.33161434531211853, + "learning_rate": 7.497851772217566e-05, + "loss": 0.0467, + "step": 22010 + }, + { + "epoch": 0.6796715846657201, + "grad_norm": 0.34157395362854004, + "learning_rate": 7.495464131639924e-05, + "loss": 0.051, + "step": 22020 + }, + { + "epoch": 0.6799802456941787, + "grad_norm": 0.3312373757362366, + "learning_rate": 7.493075733006166e-05, + "loss": 0.0476, + "step": 22030 + }, + { + "epoch": 0.6802889067226372, + "grad_norm": 0.2774185538291931, + "learning_rate": 7.490686577041828e-05, + "loss": 0.0496, + "step": 22040 + }, + { + "epoch": 0.6805975677510957, + "grad_norm": 0.2896258234977722, + "learning_rate": 7.488296664472668e-05, + "loss": 0.0494, + "step": 22050 + }, + { + "epoch": 0.6809062287795543, + "grad_norm": 0.3773074746131897, + "learning_rate": 7.485905996024682e-05, + "loss": 0.0421, + "step": 22060 + }, + { + "epoch": 0.6812148898080128, + "grad_norm": 0.6223833560943604, + "learning_rate": 7.483514572424093e-05, + "loss": 0.0493, + "step": 22070 + }, + { + "epoch": 0.6815235508364714, + "grad_norm": 0.3160802125930786, + "learning_rate": 7.481122394397349e-05, + "loss": 0.0462, + "step": 22080 + }, + { + "epoch": 0.6818322118649299, + "grad_norm": 0.31355154514312744, + "learning_rate": 7.478729462671131e-05, + "loss": 0.047, + "step": 22090 + }, + { + "epoch": 0.6821408728933884, + "grad_norm": 0.5103888511657715, + "learning_rate": 7.47633577797235e-05, + "loss": 0.0472, + "step": 22100 + }, + { + "epoch": 0.682449533921847, + "grad_norm": 0.3426433205604553, + "learning_rate": 7.473941341028144e-05, + "loss": 0.0524, + "step": 22110 + }, + { + "epoch": 0.6827581949503055, + "grad_norm": 0.3490004539489746, + "learning_rate": 7.471546152565879e-05, + "loss": 0.0463, + "step": 22120 + }, + { + "epoch": 0.6830668559787642, + "grad_norm": 0.46923503279685974, + "learning_rate": 7.46915021331315e-05, + "loss": 0.0496, + "step": 22130 + }, + { + "epoch": 0.6833755170072227, + "grad_norm": 0.3706662058830261, + "learning_rate": 7.466753523997778e-05, + "loss": 0.0462, + "step": 22140 + }, + { + "epoch": 0.6836841780356813, + "grad_norm": 0.2913872301578522, + "learning_rate": 7.464356085347819e-05, + "loss": 0.0486, + "step": 22150 + }, + { + "epoch": 0.6839928390641398, + "grad_norm": 0.39985400438308716, + "learning_rate": 7.461957898091548e-05, + "loss": 0.0515, + "step": 22160 + }, + { + "epoch": 0.6843015000925983, + "grad_norm": 0.2876930832862854, + "learning_rate": 7.459558962957473e-05, + "loss": 0.0492, + "step": 22170 + }, + { + "epoch": 0.6846101611210569, + "grad_norm": 0.31177476048469543, + "learning_rate": 7.457159280674326e-05, + "loss": 0.0486, + "step": 22180 + }, + { + "epoch": 0.6849188221495154, + "grad_norm": 0.3403491675853729, + "learning_rate": 7.454758851971066e-05, + "loss": 0.0486, + "step": 22190 + }, + { + "epoch": 0.685227483177974, + "grad_norm": 0.6301922798156738, + "learning_rate": 7.45235767757688e-05, + "loss": 0.0503, + "step": 22200 + }, + { + "epoch": 0.6855361442064325, + "grad_norm": 0.28372037410736084, + "learning_rate": 7.449955758221183e-05, + "loss": 0.0525, + "step": 22210 + }, + { + "epoch": 0.685844805234891, + "grad_norm": 0.4272341728210449, + "learning_rate": 7.447553094633615e-05, + "loss": 0.0493, + "step": 22220 + }, + { + "epoch": 0.6861534662633496, + "grad_norm": 0.4323480725288391, + "learning_rate": 7.445149687544039e-05, + "loss": 0.0473, + "step": 22230 + }, + { + "epoch": 0.6864621272918081, + "grad_norm": 0.36815372109413147, + "learning_rate": 7.44274553768255e-05, + "loss": 0.0446, + "step": 22240 + }, + { + "epoch": 0.6867707883202667, + "grad_norm": 0.342393159866333, + "learning_rate": 7.440340645779464e-05, + "loss": 0.0493, + "step": 22250 + }, + { + "epoch": 0.6870794493487252, + "grad_norm": 0.48752841353416443, + "learning_rate": 7.437935012565322e-05, + "loss": 0.0474, + "step": 22260 + }, + { + "epoch": 0.6873881103771837, + "grad_norm": 0.37264734506607056, + "learning_rate": 7.435528638770893e-05, + "loss": 0.0494, + "step": 22270 + }, + { + "epoch": 0.6876967714056423, + "grad_norm": 0.38609227538108826, + "learning_rate": 7.433121525127171e-05, + "loss": 0.0494, + "step": 22280 + }, + { + "epoch": 0.6880054324341008, + "grad_norm": 0.33378949761390686, + "learning_rate": 7.430713672365371e-05, + "loss": 0.0483, + "step": 22290 + }, + { + "epoch": 0.6883140934625595, + "grad_norm": 0.30765193700790405, + "learning_rate": 7.428305081216938e-05, + "loss": 0.0502, + "step": 22300 + }, + { + "epoch": 0.688622754491018, + "grad_norm": 0.45859578251838684, + "learning_rate": 7.425895752413536e-05, + "loss": 0.046, + "step": 22310 + }, + { + "epoch": 0.6889314155194766, + "grad_norm": 0.2358967661857605, + "learning_rate": 7.423485686687057e-05, + "loss": 0.046, + "step": 22320 + }, + { + "epoch": 0.6892400765479351, + "grad_norm": 0.32247549295425415, + "learning_rate": 7.421074884769616e-05, + "loss": 0.051, + "step": 22330 + }, + { + "epoch": 0.6895487375763936, + "grad_norm": 0.31344103813171387, + "learning_rate": 7.418663347393548e-05, + "loss": 0.0495, + "step": 22340 + }, + { + "epoch": 0.6898573986048522, + "grad_norm": 0.3073979318141937, + "learning_rate": 7.416251075291418e-05, + "loss": 0.0545, + "step": 22350 + }, + { + "epoch": 0.6901660596333107, + "grad_norm": 0.34877923130989075, + "learning_rate": 7.413838069196007e-05, + "loss": 0.045, + "step": 22360 + }, + { + "epoch": 0.6904747206617693, + "grad_norm": 0.3673745393753052, + "learning_rate": 7.411424329840324e-05, + "loss": 0.0511, + "step": 22370 + }, + { + "epoch": 0.6907833816902278, + "grad_norm": 0.319436639547348, + "learning_rate": 7.409009857957601e-05, + "loss": 0.0474, + "step": 22380 + }, + { + "epoch": 0.6910920427186863, + "grad_norm": 0.31843066215515137, + "learning_rate": 7.40659465428129e-05, + "loss": 0.0496, + "step": 22390 + }, + { + "epoch": 0.6914007037471449, + "grad_norm": 0.33809176087379456, + "learning_rate": 7.404178719545063e-05, + "loss": 0.0472, + "step": 22400 + }, + { + "epoch": 0.6917093647756034, + "grad_norm": 0.39682361483573914, + "learning_rate": 7.401762054482822e-05, + "loss": 0.0504, + "step": 22410 + }, + { + "epoch": 0.692018025804062, + "grad_norm": 0.268442839384079, + "learning_rate": 7.39934465982868e-05, + "loss": 0.0519, + "step": 22420 + }, + { + "epoch": 0.6923266868325205, + "grad_norm": 0.3369218707084656, + "learning_rate": 7.396926536316984e-05, + "loss": 0.048, + "step": 22430 + }, + { + "epoch": 0.692635347860979, + "grad_norm": 0.2670609652996063, + "learning_rate": 7.394507684682293e-05, + "loss": 0.0504, + "step": 22440 + }, + { + "epoch": 0.6929440088894376, + "grad_norm": 0.2713792026042938, + "learning_rate": 7.392088105659393e-05, + "loss": 0.0489, + "step": 22450 + }, + { + "epoch": 0.6932526699178961, + "grad_norm": 0.3702006936073303, + "learning_rate": 7.389667799983284e-05, + "loss": 0.0505, + "step": 22460 + }, + { + "epoch": 0.6935613309463547, + "grad_norm": 0.38670793175697327, + "learning_rate": 7.387246768389193e-05, + "loss": 0.046, + "step": 22470 + }, + { + "epoch": 0.6938699919748132, + "grad_norm": 0.37252479791641235, + "learning_rate": 7.384825011612563e-05, + "loss": 0.0437, + "step": 22480 + }, + { + "epoch": 0.6941786530032719, + "grad_norm": 0.3187094032764435, + "learning_rate": 7.382402530389066e-05, + "loss": 0.0411, + "step": 22490 + }, + { + "epoch": 0.6944873140317304, + "grad_norm": 0.33665987849235535, + "learning_rate": 7.379979325454582e-05, + "loss": 0.049, + "step": 22500 + }, + { + "epoch": 0.6947959750601889, + "grad_norm": 0.4310653507709503, + "learning_rate": 7.37755539754522e-05, + "loss": 0.0496, + "step": 22510 + }, + { + "epoch": 0.6951046360886475, + "grad_norm": 0.4913664162158966, + "learning_rate": 7.375130747397302e-05, + "loss": 0.0507, + "step": 22520 + }, + { + "epoch": 0.695413297117106, + "grad_norm": 0.3695093095302582, + "learning_rate": 7.372705375747377e-05, + "loss": 0.0432, + "step": 22530 + }, + { + "epoch": 0.6957219581455646, + "grad_norm": 0.3851015269756317, + "learning_rate": 7.370279283332205e-05, + "loss": 0.051, + "step": 22540 + }, + { + "epoch": 0.6960306191740231, + "grad_norm": 0.30453023314476013, + "learning_rate": 7.36785247088877e-05, + "loss": 0.0474, + "step": 22550 + }, + { + "epoch": 0.6963392802024816, + "grad_norm": 0.3848097026348114, + "learning_rate": 7.365424939154275e-05, + "loss": 0.0436, + "step": 22560 + }, + { + "epoch": 0.6966479412309402, + "grad_norm": 0.48171722888946533, + "learning_rate": 7.362996688866138e-05, + "loss": 0.048, + "step": 22570 + }, + { + "epoch": 0.6969566022593987, + "grad_norm": 0.4030483067035675, + "learning_rate": 7.360567720761999e-05, + "loss": 0.0489, + "step": 22580 + }, + { + "epoch": 0.6972652632878573, + "grad_norm": 0.3324512839317322, + "learning_rate": 7.358138035579711e-05, + "loss": 0.0458, + "step": 22590 + }, + { + "epoch": 0.6975739243163158, + "grad_norm": 0.372232586145401, + "learning_rate": 7.355707634057354e-05, + "loss": 0.0463, + "step": 22600 + }, + { + "epoch": 0.6978825853447743, + "grad_norm": 0.2437884360551834, + "learning_rate": 7.353276516933215e-05, + "loss": 0.0505, + "step": 22610 + }, + { + "epoch": 0.6981912463732329, + "grad_norm": 0.3062092363834381, + "learning_rate": 7.350844684945806e-05, + "loss": 0.0492, + "step": 22620 + }, + { + "epoch": 0.6984999074016914, + "grad_norm": 0.28370368480682373, + "learning_rate": 7.348412138833851e-05, + "loss": 0.0482, + "step": 22630 + }, + { + "epoch": 0.69880856843015, + "grad_norm": 0.3824000060558319, + "learning_rate": 7.345978879336295e-05, + "loss": 0.0448, + "step": 22640 + }, + { + "epoch": 0.6991172294586085, + "grad_norm": 0.48266273736953735, + "learning_rate": 7.343544907192296e-05, + "loss": 0.0478, + "step": 22650 + }, + { + "epoch": 0.699425890487067, + "grad_norm": 0.5465936660766602, + "learning_rate": 7.341110223141235e-05, + "loss": 0.0473, + "step": 22660 + }, + { + "epoch": 0.6997345515155257, + "grad_norm": 0.28734922409057617, + "learning_rate": 7.3386748279227e-05, + "loss": 0.0489, + "step": 22670 + }, + { + "epoch": 0.7000432125439842, + "grad_norm": 0.37404653429985046, + "learning_rate": 7.336238722276501e-05, + "loss": 0.0478, + "step": 22680 + }, + { + "epoch": 0.7003518735724428, + "grad_norm": 0.3623608946800232, + "learning_rate": 7.333801906942663e-05, + "loss": 0.0506, + "step": 22690 + }, + { + "epoch": 0.7006605346009013, + "grad_norm": 0.39533528685569763, + "learning_rate": 7.331364382661428e-05, + "loss": 0.0519, + "step": 22700 + }, + { + "epoch": 0.7009691956293599, + "grad_norm": 0.35384270548820496, + "learning_rate": 7.328926150173248e-05, + "loss": 0.0521, + "step": 22710 + }, + { + "epoch": 0.7012778566578184, + "grad_norm": 0.27482783794403076, + "learning_rate": 7.326487210218795e-05, + "loss": 0.0535, + "step": 22720 + }, + { + "epoch": 0.7015865176862769, + "grad_norm": 0.3139359951019287, + "learning_rate": 7.324047563538955e-05, + "loss": 0.0426, + "step": 22730 + }, + { + "epoch": 0.7018951787147355, + "grad_norm": 0.41461965441703796, + "learning_rate": 7.321607210874828e-05, + "loss": 0.0497, + "step": 22740 + }, + { + "epoch": 0.702203839743194, + "grad_norm": 0.3619934320449829, + "learning_rate": 7.31916615296773e-05, + "loss": 0.0488, + "step": 22750 + }, + { + "epoch": 0.7025125007716526, + "grad_norm": 0.3854247033596039, + "learning_rate": 7.316724390559188e-05, + "loss": 0.0483, + "step": 22760 + }, + { + "epoch": 0.7028211618001111, + "grad_norm": 0.28127220273017883, + "learning_rate": 7.314281924390946e-05, + "loss": 0.0474, + "step": 22770 + }, + { + "epoch": 0.7031298228285696, + "grad_norm": 0.3175477981567383, + "learning_rate": 7.311838755204959e-05, + "loss": 0.0505, + "step": 22780 + }, + { + "epoch": 0.7034384838570282, + "grad_norm": 0.3770820200443268, + "learning_rate": 7.3093948837434e-05, + "loss": 0.0505, + "step": 22790 + }, + { + "epoch": 0.7037471448854867, + "grad_norm": 0.3982820510864258, + "learning_rate": 7.306950310748651e-05, + "loss": 0.0444, + "step": 22800 + }, + { + "epoch": 0.7040558059139453, + "grad_norm": 0.37992212176322937, + "learning_rate": 7.304505036963311e-05, + "loss": 0.0492, + "step": 22810 + }, + { + "epoch": 0.7043644669424038, + "grad_norm": 0.278139591217041, + "learning_rate": 7.302059063130186e-05, + "loss": 0.0431, + "step": 22820 + }, + { + "epoch": 0.7046731279708623, + "grad_norm": 0.31977346539497375, + "learning_rate": 7.2996123899923e-05, + "loss": 0.0449, + "step": 22830 + }, + { + "epoch": 0.704981788999321, + "grad_norm": 0.4060536026954651, + "learning_rate": 7.297165018292886e-05, + "loss": 0.0473, + "step": 22840 + }, + { + "epoch": 0.7052904500277795, + "grad_norm": 0.23268026113510132, + "learning_rate": 7.294716948775396e-05, + "loss": 0.0426, + "step": 22850 + }, + { + "epoch": 0.7055991110562381, + "grad_norm": 0.35000476241111755, + "learning_rate": 7.292268182183484e-05, + "loss": 0.0484, + "step": 22860 + }, + { + "epoch": 0.7059077720846966, + "grad_norm": 0.5210880637168884, + "learning_rate": 7.28981871926102e-05, + "loss": 0.0507, + "step": 22870 + }, + { + "epoch": 0.7062164331131552, + "grad_norm": 0.3219224214553833, + "learning_rate": 7.28736856075209e-05, + "loss": 0.0485, + "step": 22880 + }, + { + "epoch": 0.7065250941416137, + "grad_norm": 0.3703272044658661, + "learning_rate": 7.284917707400985e-05, + "loss": 0.0439, + "step": 22890 + }, + { + "epoch": 0.7068337551700722, + "grad_norm": 0.328532338142395, + "learning_rate": 7.282466159952212e-05, + "loss": 0.0444, + "step": 22900 + }, + { + "epoch": 0.7071424161985308, + "grad_norm": 0.5862252116203308, + "learning_rate": 7.280013919150483e-05, + "loss": 0.0481, + "step": 22910 + }, + { + "epoch": 0.7074510772269893, + "grad_norm": 0.3250638544559479, + "learning_rate": 7.277560985740728e-05, + "loss": 0.0513, + "step": 22920 + }, + { + "epoch": 0.7077597382554479, + "grad_norm": 0.3371751606464386, + "learning_rate": 7.275107360468079e-05, + "loss": 0.0471, + "step": 22930 + }, + { + "epoch": 0.7080683992839064, + "grad_norm": 0.6876505613327026, + "learning_rate": 7.272653044077885e-05, + "loss": 0.0508, + "step": 22940 + }, + { + "epoch": 0.7083770603123649, + "grad_norm": 0.40435081720352173, + "learning_rate": 7.270198037315703e-05, + "loss": 0.0445, + "step": 22950 + }, + { + "epoch": 0.7086857213408235, + "grad_norm": 0.40607109665870667, + "learning_rate": 7.267742340927297e-05, + "loss": 0.0452, + "step": 22960 + }, + { + "epoch": 0.708994382369282, + "grad_norm": 0.3659418821334839, + "learning_rate": 7.265285955658645e-05, + "loss": 0.0463, + "step": 22970 + }, + { + "epoch": 0.7093030433977406, + "grad_norm": 0.39504626393318176, + "learning_rate": 7.26282888225593e-05, + "loss": 0.0476, + "step": 22980 + }, + { + "epoch": 0.7096117044261991, + "grad_norm": 0.4344708323478699, + "learning_rate": 7.260371121465548e-05, + "loss": 0.049, + "step": 22990 + }, + { + "epoch": 0.7099203654546576, + "grad_norm": 0.35964611172676086, + "learning_rate": 7.2579126740341e-05, + "loss": 0.0454, + "step": 23000 + }, + { + "epoch": 0.7102290264831163, + "grad_norm": 0.43665826320648193, + "learning_rate": 7.2554535407084e-05, + "loss": 0.0486, + "step": 23010 + }, + { + "epoch": 0.7105376875115748, + "grad_norm": 0.2968241572380066, + "learning_rate": 7.252993722235464e-05, + "loss": 0.0442, + "step": 23020 + }, + { + "epoch": 0.7108463485400334, + "grad_norm": 0.412322074174881, + "learning_rate": 7.250533219362523e-05, + "loss": 0.0488, + "step": 23030 + }, + { + "epoch": 0.7111550095684919, + "grad_norm": 0.305276483297348, + "learning_rate": 7.248072032837012e-05, + "loss": 0.0455, + "step": 23040 + }, + { + "epoch": 0.7114636705969505, + "grad_norm": 0.3178154528141022, + "learning_rate": 7.245610163406575e-05, + "loss": 0.0553, + "step": 23050 + }, + { + "epoch": 0.711772331625409, + "grad_norm": 0.27704960107803345, + "learning_rate": 7.243147611819061e-05, + "loss": 0.0455, + "step": 23060 + }, + { + "epoch": 0.7120809926538675, + "grad_norm": 0.41484782099723816, + "learning_rate": 7.240684378822531e-05, + "loss": 0.0425, + "step": 23070 + }, + { + "epoch": 0.7123896536823261, + "grad_norm": 0.40933021903038025, + "learning_rate": 7.238220465165248e-05, + "loss": 0.047, + "step": 23080 + }, + { + "epoch": 0.7126983147107846, + "grad_norm": 0.38098523020744324, + "learning_rate": 7.235755871595684e-05, + "loss": 0.0458, + "step": 23090 + }, + { + "epoch": 0.7130069757392432, + "grad_norm": 0.3075089752674103, + "learning_rate": 7.233290598862517e-05, + "loss": 0.0495, + "step": 23100 + }, + { + "epoch": 0.7133156367677017, + "grad_norm": 0.29166996479034424, + "learning_rate": 7.230824647714635e-05, + "loss": 0.0511, + "step": 23110 + }, + { + "epoch": 0.7136242977961602, + "grad_norm": 0.3639003038406372, + "learning_rate": 7.228358018901124e-05, + "loss": 0.049, + "step": 23120 + }, + { + "epoch": 0.7139329588246188, + "grad_norm": 0.348391056060791, + "learning_rate": 7.225890713171286e-05, + "loss": 0.0466, + "step": 23130 + }, + { + "epoch": 0.7142416198530773, + "grad_norm": 0.3611622750759125, + "learning_rate": 7.223422731274618e-05, + "loss": 0.0436, + "step": 23140 + }, + { + "epoch": 0.7145502808815359, + "grad_norm": 0.2994506359100342, + "learning_rate": 7.220954073960832e-05, + "loss": 0.0507, + "step": 23150 + }, + { + "epoch": 0.7148589419099944, + "grad_norm": 0.34931397438049316, + "learning_rate": 7.218484741979838e-05, + "loss": 0.0498, + "step": 23160 + }, + { + "epoch": 0.7151676029384529, + "grad_norm": 0.3311139941215515, + "learning_rate": 7.216014736081756e-05, + "loss": 0.0488, + "step": 23170 + }, + { + "epoch": 0.7154762639669116, + "grad_norm": 0.2962479889392853, + "learning_rate": 7.213544057016906e-05, + "loss": 0.0506, + "step": 23180 + }, + { + "epoch": 0.71578492499537, + "grad_norm": 0.3475821018218994, + "learning_rate": 7.211072705535819e-05, + "loss": 0.0499, + "step": 23190 + }, + { + "epoch": 0.7160935860238287, + "grad_norm": 0.34937790036201477, + "learning_rate": 7.208600682389224e-05, + "loss": 0.0461, + "step": 23200 + }, + { + "epoch": 0.7164022470522872, + "grad_norm": 0.3494119346141815, + "learning_rate": 7.206127988328055e-05, + "loss": 0.0435, + "step": 23210 + }, + { + "epoch": 0.7167109080807458, + "grad_norm": 0.32163864374160767, + "learning_rate": 7.203654624103453e-05, + "loss": 0.0465, + "step": 23220 + }, + { + "epoch": 0.7170195691092043, + "grad_norm": 0.5778687000274658, + "learning_rate": 7.201180590466761e-05, + "loss": 0.0472, + "step": 23230 + }, + { + "epoch": 0.7173282301376628, + "grad_norm": 0.23982253670692444, + "learning_rate": 7.198705888169523e-05, + "loss": 0.0528, + "step": 23240 + }, + { + "epoch": 0.7176368911661214, + "grad_norm": 0.4301946461200714, + "learning_rate": 7.196230517963491e-05, + "loss": 0.0459, + "step": 23250 + }, + { + "epoch": 0.7179455521945799, + "grad_norm": 0.4720837473869324, + "learning_rate": 7.193754480600615e-05, + "loss": 0.0515, + "step": 23260 + }, + { + "epoch": 0.7182542132230385, + "grad_norm": 0.4326397180557251, + "learning_rate": 7.19127777683305e-05, + "loss": 0.0486, + "step": 23270 + }, + { + "epoch": 0.718562874251497, + "grad_norm": 0.33960142731666565, + "learning_rate": 7.188800407413156e-05, + "loss": 0.0482, + "step": 23280 + }, + { + "epoch": 0.7188715352799555, + "grad_norm": 0.5723434686660767, + "learning_rate": 7.186322373093489e-05, + "loss": 0.0444, + "step": 23290 + }, + { + "epoch": 0.7191801963084141, + "grad_norm": 0.39138779044151306, + "learning_rate": 7.18384367462681e-05, + "loss": 0.0448, + "step": 23300 + }, + { + "epoch": 0.7194888573368726, + "grad_norm": 0.3226468563079834, + "learning_rate": 7.181364312766085e-05, + "loss": 0.0491, + "step": 23310 + }, + { + "epoch": 0.7197975183653312, + "grad_norm": 0.36503472924232483, + "learning_rate": 7.178884288264477e-05, + "loss": 0.0442, + "step": 23320 + }, + { + "epoch": 0.7201061793937897, + "grad_norm": 0.31538158655166626, + "learning_rate": 7.176403601875353e-05, + "loss": 0.0427, + "step": 23330 + }, + { + "epoch": 0.7204148404222482, + "grad_norm": 0.45673367381095886, + "learning_rate": 7.173922254352279e-05, + "loss": 0.0476, + "step": 23340 + }, + { + "epoch": 0.7207235014507068, + "grad_norm": 0.3171766698360443, + "learning_rate": 7.171440246449024e-05, + "loss": 0.047, + "step": 23350 + }, + { + "epoch": 0.7210321624791654, + "grad_norm": 0.27468740940093994, + "learning_rate": 7.168957578919555e-05, + "loss": 0.0485, + "step": 23360 + }, + { + "epoch": 0.721340823507624, + "grad_norm": 0.30955788493156433, + "learning_rate": 7.16647425251804e-05, + "loss": 0.0499, + "step": 23370 + }, + { + "epoch": 0.7216494845360825, + "grad_norm": 0.3698763847351074, + "learning_rate": 7.163990267998852e-05, + "loss": 0.0479, + "step": 23380 + }, + { + "epoch": 0.7219581455645411, + "grad_norm": 0.35705819725990295, + "learning_rate": 7.161505626116556e-05, + "loss": 0.0462, + "step": 23390 + }, + { + "epoch": 0.7222668065929996, + "grad_norm": 0.27314135432243347, + "learning_rate": 7.159020327625923e-05, + "loss": 0.0463, + "step": 23400 + }, + { + "epoch": 0.7225754676214581, + "grad_norm": 0.24950474500656128, + "learning_rate": 7.15653437328192e-05, + "loss": 0.0475, + "step": 23410 + }, + { + "epoch": 0.7228841286499167, + "grad_norm": 0.29120394587516785, + "learning_rate": 7.154047763839713e-05, + "loss": 0.0478, + "step": 23420 + }, + { + "epoch": 0.7231927896783752, + "grad_norm": 0.4636458158493042, + "learning_rate": 7.15156050005467e-05, + "loss": 0.0492, + "step": 23430 + }, + { + "epoch": 0.7235014507068338, + "grad_norm": 0.36814600229263306, + "learning_rate": 7.149072582682357e-05, + "loss": 0.0453, + "step": 23440 + }, + { + "epoch": 0.7238101117352923, + "grad_norm": 0.337497353553772, + "learning_rate": 7.146584012478535e-05, + "loss": 0.0439, + "step": 23450 + }, + { + "epoch": 0.7241187727637508, + "grad_norm": 0.35315680503845215, + "learning_rate": 7.144094790199169e-05, + "loss": 0.0471, + "step": 23460 + }, + { + "epoch": 0.7244274337922094, + "grad_norm": 0.508025586605072, + "learning_rate": 7.141604916600415e-05, + "loss": 0.0509, + "step": 23470 + }, + { + "epoch": 0.7247360948206679, + "grad_norm": 0.3122338056564331, + "learning_rate": 7.139114392438635e-05, + "loss": 0.0504, + "step": 23480 + }, + { + "epoch": 0.7250447558491265, + "grad_norm": 0.3780355155467987, + "learning_rate": 7.136623218470382e-05, + "loss": 0.0425, + "step": 23490 + }, + { + "epoch": 0.725353416877585, + "grad_norm": 0.3727096915245056, + "learning_rate": 7.13413139545241e-05, + "loss": 0.0478, + "step": 23500 + }, + { + "epoch": 0.7256620779060435, + "grad_norm": 0.30377933382987976, + "learning_rate": 7.131638924141668e-05, + "loss": 0.049, + "step": 23510 + }, + { + "epoch": 0.7259707389345021, + "grad_norm": 0.28336676955223083, + "learning_rate": 7.129145805295304e-05, + "loss": 0.0487, + "step": 23520 + }, + { + "epoch": 0.7262793999629606, + "grad_norm": 0.4014594256877899, + "learning_rate": 7.126652039670661e-05, + "loss": 0.0476, + "step": 23530 + }, + { + "epoch": 0.7265880609914193, + "grad_norm": 0.31013524532318115, + "learning_rate": 7.124157628025278e-05, + "loss": 0.0499, + "step": 23540 + }, + { + "epoch": 0.7268967220198778, + "grad_norm": 0.3332870304584503, + "learning_rate": 7.121662571116894e-05, + "loss": 0.0489, + "step": 23550 + }, + { + "epoch": 0.7272053830483363, + "grad_norm": 0.401683509349823, + "learning_rate": 7.119166869703441e-05, + "loss": 0.0486, + "step": 23560 + }, + { + "epoch": 0.7275140440767949, + "grad_norm": 0.30997341871261597, + "learning_rate": 7.116670524543044e-05, + "loss": 0.0483, + "step": 23570 + }, + { + "epoch": 0.7278227051052534, + "grad_norm": 0.520528256893158, + "learning_rate": 7.114173536394032e-05, + "loss": 0.0461, + "step": 23580 + }, + { + "epoch": 0.728131366133712, + "grad_norm": 0.3036428987979889, + "learning_rate": 7.111675906014917e-05, + "loss": 0.045, + "step": 23590 + }, + { + "epoch": 0.7284400271621705, + "grad_norm": 0.364646852016449, + "learning_rate": 7.109177634164421e-05, + "loss": 0.0474, + "step": 23600 + }, + { + "epoch": 0.7287486881906291, + "grad_norm": 0.3455749750137329, + "learning_rate": 7.106678721601449e-05, + "loss": 0.0466, + "step": 23610 + }, + { + "epoch": 0.7290573492190876, + "grad_norm": 0.3013343811035156, + "learning_rate": 7.104179169085103e-05, + "loss": 0.0483, + "step": 23620 + }, + { + "epoch": 0.7293660102475461, + "grad_norm": 0.3454049825668335, + "learning_rate": 7.101678977374683e-05, + "loss": 0.0462, + "step": 23630 + }, + { + "epoch": 0.7296746712760047, + "grad_norm": 0.2885112166404724, + "learning_rate": 7.099178147229685e-05, + "loss": 0.0433, + "step": 23640 + }, + { + "epoch": 0.7299833323044632, + "grad_norm": 0.42617326974868774, + "learning_rate": 7.096676679409789e-05, + "loss": 0.0489, + "step": 23650 + }, + { + "epoch": 0.7302919933329218, + "grad_norm": 0.3860354423522949, + "learning_rate": 7.094174574674877e-05, + "loss": 0.0538, + "step": 23660 + }, + { + "epoch": 0.7306006543613803, + "grad_norm": 0.3416774868965149, + "learning_rate": 7.091671833785025e-05, + "loss": 0.047, + "step": 23670 + }, + { + "epoch": 0.7309093153898388, + "grad_norm": 0.4106703996658325, + "learning_rate": 7.089168457500493e-05, + "loss": 0.0439, + "step": 23680 + }, + { + "epoch": 0.7312179764182974, + "grad_norm": 0.2828470766544342, + "learning_rate": 7.086664446581747e-05, + "loss": 0.0474, + "step": 23690 + }, + { + "epoch": 0.731526637446756, + "grad_norm": 0.4841563403606415, + "learning_rate": 7.084159801789438e-05, + "loss": 0.0463, + "step": 23700 + }, + { + "epoch": 0.7318352984752146, + "grad_norm": 0.36759260296821594, + "learning_rate": 7.081654523884411e-05, + "loss": 0.0488, + "step": 23710 + }, + { + "epoch": 0.7321439595036731, + "grad_norm": 0.2683006525039673, + "learning_rate": 7.0791486136277e-05, + "loss": 0.0477, + "step": 23720 + }, + { + "epoch": 0.7324526205321316, + "grad_norm": 0.35363879799842834, + "learning_rate": 7.07664207178054e-05, + "loss": 0.0456, + "step": 23730 + }, + { + "epoch": 0.7327612815605902, + "grad_norm": 0.2748877704143524, + "learning_rate": 7.074134899104345e-05, + "loss": 0.0478, + "step": 23740 + }, + { + "epoch": 0.7330699425890487, + "grad_norm": 0.329047292470932, + "learning_rate": 7.071627096360735e-05, + "loss": 0.0481, + "step": 23750 + }, + { + "epoch": 0.7333786036175073, + "grad_norm": 0.32160425186157227, + "learning_rate": 7.069118664311511e-05, + "loss": 0.0484, + "step": 23760 + }, + { + "epoch": 0.7336872646459658, + "grad_norm": 0.3088441491127014, + "learning_rate": 7.06660960371867e-05, + "loss": 0.0455, + "step": 23770 + }, + { + "epoch": 0.7339959256744244, + "grad_norm": 0.32612791657447815, + "learning_rate": 7.064099915344396e-05, + "loss": 0.048, + "step": 23780 + }, + { + "epoch": 0.7343045867028829, + "grad_norm": 0.36274614930152893, + "learning_rate": 7.061589599951066e-05, + "loss": 0.048, + "step": 23790 + }, + { + "epoch": 0.7346132477313414, + "grad_norm": 0.3219137191772461, + "learning_rate": 7.05907865830125e-05, + "loss": 0.0476, + "step": 23800 + }, + { + "epoch": 0.7349219087598, + "grad_norm": 0.3364027142524719, + "learning_rate": 7.056567091157703e-05, + "loss": 0.0504, + "step": 23810 + }, + { + "epoch": 0.7352305697882585, + "grad_norm": 0.4719105362892151, + "learning_rate": 7.054054899283375e-05, + "loss": 0.0503, + "step": 23820 + }, + { + "epoch": 0.7355392308167171, + "grad_norm": 0.31685832142829895, + "learning_rate": 7.051542083441403e-05, + "loss": 0.0517, + "step": 23830 + }, + { + "epoch": 0.7358478918451756, + "grad_norm": 0.31916967034339905, + "learning_rate": 7.049028644395113e-05, + "loss": 0.0478, + "step": 23840 + }, + { + "epoch": 0.7361565528736341, + "grad_norm": 0.29986926913261414, + "learning_rate": 7.046514582908024e-05, + "loss": 0.0512, + "step": 23850 + }, + { + "epoch": 0.7364652139020927, + "grad_norm": 0.24025951325893402, + "learning_rate": 7.043999899743838e-05, + "loss": 0.0461, + "step": 23860 + }, + { + "epoch": 0.7367738749305512, + "grad_norm": 0.3913503587245941, + "learning_rate": 7.041484595666451e-05, + "loss": 0.0483, + "step": 23870 + }, + { + "epoch": 0.7370825359590099, + "grad_norm": 0.33127617835998535, + "learning_rate": 7.038968671439948e-05, + "loss": 0.0452, + "step": 23880 + }, + { + "epoch": 0.7373911969874684, + "grad_norm": 0.28130942583084106, + "learning_rate": 7.036452127828596e-05, + "loss": 0.046, + "step": 23890 + }, + { + "epoch": 0.7376998580159269, + "grad_norm": 0.30850809812545776, + "learning_rate": 7.033934965596859e-05, + "loss": 0.0529, + "step": 23900 + }, + { + "epoch": 0.7380085190443855, + "grad_norm": 0.4325932562351227, + "learning_rate": 7.031417185509381e-05, + "loss": 0.0494, + "step": 23910 + }, + { + "epoch": 0.738317180072844, + "grad_norm": 0.3679579794406891, + "learning_rate": 7.028898788331e-05, + "loss": 0.0427, + "step": 23920 + }, + { + "epoch": 0.7386258411013026, + "grad_norm": 0.3611748516559601, + "learning_rate": 7.026379774826736e-05, + "loss": 0.0458, + "step": 23930 + }, + { + "epoch": 0.7389345021297611, + "grad_norm": 0.27864354848861694, + "learning_rate": 7.0238601457618e-05, + "loss": 0.0481, + "step": 23940 + }, + { + "epoch": 0.7392431631582197, + "grad_norm": 0.9952681660652161, + "learning_rate": 7.02133990190159e-05, + "loss": 0.048, + "step": 23950 + }, + { + "epoch": 0.7395518241866782, + "grad_norm": 0.3791653513908386, + "learning_rate": 7.018819044011687e-05, + "loss": 0.0524, + "step": 23960 + }, + { + "epoch": 0.7398604852151367, + "grad_norm": 0.4050484895706177, + "learning_rate": 7.016297572857863e-05, + "loss": 0.0488, + "step": 23970 + }, + { + "epoch": 0.7401691462435953, + "grad_norm": 0.5192070603370667, + "learning_rate": 7.013775489206072e-05, + "loss": 0.0476, + "step": 23980 + }, + { + "epoch": 0.7404778072720538, + "grad_norm": 0.3864498734474182, + "learning_rate": 7.01125279382246e-05, + "loss": 0.044, + "step": 23990 + }, + { + "epoch": 0.7407864683005124, + "grad_norm": 0.3869195878505707, + "learning_rate": 7.008729487473351e-05, + "loss": 0.0434, + "step": 24000 + }, + { + "epoch": 0.7410951293289709, + "grad_norm": 0.3022224009037018, + "learning_rate": 7.006205570925263e-05, + "loss": 0.048, + "step": 24010 + }, + { + "epoch": 0.7414037903574294, + "grad_norm": 0.4578932225704193, + "learning_rate": 7.003681044944892e-05, + "loss": 0.0474, + "step": 24020 + }, + { + "epoch": 0.741712451385888, + "grad_norm": 0.30986565351486206, + "learning_rate": 7.001155910299126e-05, + "loss": 0.0459, + "step": 24030 + }, + { + "epoch": 0.7420211124143465, + "grad_norm": 0.3835034966468811, + "learning_rate": 6.99863016775503e-05, + "loss": 0.0448, + "step": 24040 + }, + { + "epoch": 0.7423297734428052, + "grad_norm": 0.44651323556900024, + "learning_rate": 6.996103818079859e-05, + "loss": 0.0462, + "step": 24050 + }, + { + "epoch": 0.7426384344712637, + "grad_norm": 0.2837642431259155, + "learning_rate": 6.993576862041054e-05, + "loss": 0.0458, + "step": 24060 + }, + { + "epoch": 0.7429470954997222, + "grad_norm": 0.37228962779045105, + "learning_rate": 6.991049300406235e-05, + "loss": 0.0453, + "step": 24070 + }, + { + "epoch": 0.7432557565281808, + "grad_norm": 0.38277196884155273, + "learning_rate": 6.988521133943209e-05, + "loss": 0.0482, + "step": 24080 + }, + { + "epoch": 0.7435644175566393, + "grad_norm": 0.32490262389183044, + "learning_rate": 6.985992363419966e-05, + "loss": 0.0454, + "step": 24090 + }, + { + "epoch": 0.7438730785850979, + "grad_norm": 0.5400202870368958, + "learning_rate": 6.983462989604682e-05, + "loss": 0.0506, + "step": 24100 + }, + { + "epoch": 0.7441817396135564, + "grad_norm": 0.29486197233200073, + "learning_rate": 6.980933013265709e-05, + "loss": 0.0475, + "step": 24110 + }, + { + "epoch": 0.744490400642015, + "grad_norm": 0.3050416111946106, + "learning_rate": 6.978402435171592e-05, + "loss": 0.045, + "step": 24120 + }, + { + "epoch": 0.7447990616704735, + "grad_norm": 0.38706454634666443, + "learning_rate": 6.975871256091052e-05, + "loss": 0.0491, + "step": 24130 + }, + { + "epoch": 0.745107722698932, + "grad_norm": 0.4743967056274414, + "learning_rate": 6.973339476792995e-05, + "loss": 0.0469, + "step": 24140 + }, + { + "epoch": 0.7454163837273906, + "grad_norm": 0.5151547193527222, + "learning_rate": 6.970807098046505e-05, + "loss": 0.0441, + "step": 24150 + }, + { + "epoch": 0.7457250447558491, + "grad_norm": 0.40524452924728394, + "learning_rate": 6.968274120620858e-05, + "loss": 0.0471, + "step": 24160 + }, + { + "epoch": 0.7460337057843077, + "grad_norm": 0.42615780234336853, + "learning_rate": 6.965740545285499e-05, + "loss": 0.0499, + "step": 24170 + }, + { + "epoch": 0.7463423668127662, + "grad_norm": 0.6229512095451355, + "learning_rate": 6.963206372810068e-05, + "loss": 0.0491, + "step": 24180 + }, + { + "epoch": 0.7466510278412247, + "grad_norm": 0.3979022204875946, + "learning_rate": 6.960671603964375e-05, + "loss": 0.0398, + "step": 24190 + }, + { + "epoch": 0.7469596888696833, + "grad_norm": 0.3917567729949951, + "learning_rate": 6.958136239518418e-05, + "loss": 0.0478, + "step": 24200 + }, + { + "epoch": 0.7472683498981418, + "grad_norm": 0.32907193899154663, + "learning_rate": 6.955600280242371e-05, + "loss": 0.0494, + "step": 24210 + }, + { + "epoch": 0.7475770109266004, + "grad_norm": 0.41656413674354553, + "learning_rate": 6.953063726906596e-05, + "loss": 0.0473, + "step": 24220 + }, + { + "epoch": 0.747885671955059, + "grad_norm": 0.5845019817352295, + "learning_rate": 6.950526580281626e-05, + "loss": 0.0441, + "step": 24230 + }, + { + "epoch": 0.7481943329835175, + "grad_norm": 0.36986052989959717, + "learning_rate": 6.947988841138184e-05, + "loss": 0.0445, + "step": 24240 + }, + { + "epoch": 0.7485029940119761, + "grad_norm": 0.38757166266441345, + "learning_rate": 6.945450510247165e-05, + "loss": 0.047, + "step": 24250 + }, + { + "epoch": 0.7488116550404346, + "grad_norm": 0.4309564530849457, + "learning_rate": 6.942911588379647e-05, + "loss": 0.0447, + "step": 24260 + }, + { + "epoch": 0.7491203160688932, + "grad_norm": 0.37694570422172546, + "learning_rate": 6.940372076306888e-05, + "loss": 0.0439, + "step": 24270 + }, + { + "epoch": 0.7494289770973517, + "grad_norm": 0.33615821599960327, + "learning_rate": 6.937831974800326e-05, + "loss": 0.0457, + "step": 24280 + }, + { + "epoch": 0.7497376381258103, + "grad_norm": 0.4111194610595703, + "learning_rate": 6.935291284631574e-05, + "loss": 0.0502, + "step": 24290 + }, + { + "epoch": 0.7500462991542688, + "grad_norm": 0.3064897656440735, + "learning_rate": 6.932750006572428e-05, + "loss": 0.0505, + "step": 24300 + }, + { + "epoch": 0.7503549601827273, + "grad_norm": 0.3891907036304474, + "learning_rate": 6.930208141394863e-05, + "loss": 0.0506, + "step": 24310 + }, + { + "epoch": 0.7506636212111859, + "grad_norm": 0.4106622636318207, + "learning_rate": 6.927665689871026e-05, + "loss": 0.0468, + "step": 24320 + }, + { + "epoch": 0.7509722822396444, + "grad_norm": 0.4955662190914154, + "learning_rate": 6.925122652773253e-05, + "loss": 0.0455, + "step": 24330 + }, + { + "epoch": 0.751280943268103, + "grad_norm": 0.2659757137298584, + "learning_rate": 6.922579030874046e-05, + "loss": 0.0491, + "step": 24340 + }, + { + "epoch": 0.7515896042965615, + "grad_norm": 0.43829500675201416, + "learning_rate": 6.920034824946093e-05, + "loss": 0.0487, + "step": 24350 + }, + { + "epoch": 0.75189826532502, + "grad_norm": 0.3700326681137085, + "learning_rate": 6.917490035762255e-05, + "loss": 0.0453, + "step": 24360 + }, + { + "epoch": 0.7522069263534786, + "grad_norm": 0.3888901174068451, + "learning_rate": 6.914944664095573e-05, + "loss": 0.0483, + "step": 24370 + }, + { + "epoch": 0.7525155873819371, + "grad_norm": 0.35021233558654785, + "learning_rate": 6.912398710719264e-05, + "loss": 0.0461, + "step": 24380 + }, + { + "epoch": 0.7528242484103957, + "grad_norm": 0.3289701044559479, + "learning_rate": 6.90985217640672e-05, + "loss": 0.0455, + "step": 24390 + }, + { + "epoch": 0.7531329094388542, + "grad_norm": 0.5529247522354126, + "learning_rate": 6.90730506193151e-05, + "loss": 0.0501, + "step": 24400 + }, + { + "epoch": 0.7534415704673127, + "grad_norm": 0.37052619457244873, + "learning_rate": 6.904757368067384e-05, + "loss": 0.0518, + "step": 24410 + }, + { + "epoch": 0.7537502314957714, + "grad_norm": 0.2949816584587097, + "learning_rate": 6.90220909558826e-05, + "loss": 0.0432, + "step": 24420 + }, + { + "epoch": 0.7540588925242299, + "grad_norm": 0.39759591221809387, + "learning_rate": 6.899660245268237e-05, + "loss": 0.05, + "step": 24430 + }, + { + "epoch": 0.7543675535526885, + "grad_norm": 0.3203888237476349, + "learning_rate": 6.897110817881592e-05, + "loss": 0.0492, + "step": 24440 + }, + { + "epoch": 0.754676214581147, + "grad_norm": 0.2584962546825409, + "learning_rate": 6.894560814202769e-05, + "loss": 0.0506, + "step": 24450 + }, + { + "epoch": 0.7549848756096055, + "grad_norm": 0.3655581474304199, + "learning_rate": 6.892010235006394e-05, + "loss": 0.0422, + "step": 24460 + }, + { + "epoch": 0.7552935366380641, + "grad_norm": 0.4174065887928009, + "learning_rate": 6.889459081067264e-05, + "loss": 0.0435, + "step": 24470 + }, + { + "epoch": 0.7556021976665226, + "grad_norm": 0.32798677682876587, + "learning_rate": 6.886907353160356e-05, + "loss": 0.0509, + "step": 24480 + }, + { + "epoch": 0.7559108586949812, + "grad_norm": 0.42029157280921936, + "learning_rate": 6.884355052060814e-05, + "loss": 0.0445, + "step": 24490 + }, + { + "epoch": 0.7562195197234397, + "grad_norm": 0.3531777858734131, + "learning_rate": 6.88180217854396e-05, + "loss": 0.0451, + "step": 24500 + }, + { + "epoch": 0.7565281807518983, + "grad_norm": 0.4223526120185852, + "learning_rate": 6.87924873338529e-05, + "loss": 0.0438, + "step": 24510 + }, + { + "epoch": 0.7568368417803568, + "grad_norm": 0.30598604679107666, + "learning_rate": 6.876694717360475e-05, + "loss": 0.0444, + "step": 24520 + }, + { + "epoch": 0.7571455028088153, + "grad_norm": 0.2907758057117462, + "learning_rate": 6.874140131245355e-05, + "loss": 0.0457, + "step": 24530 + }, + { + "epoch": 0.7574541638372739, + "grad_norm": 0.3484226167201996, + "learning_rate": 6.871584975815948e-05, + "loss": 0.0461, + "step": 24540 + }, + { + "epoch": 0.7577628248657324, + "grad_norm": 0.5151646733283997, + "learning_rate": 6.86902925184844e-05, + "loss": 0.0459, + "step": 24550 + }, + { + "epoch": 0.758071485894191, + "grad_norm": 0.3795390725135803, + "learning_rate": 6.866472960119195e-05, + "loss": 0.0466, + "step": 24560 + }, + { + "epoch": 0.7583801469226495, + "grad_norm": 0.40297847986221313, + "learning_rate": 6.863916101404748e-05, + "loss": 0.0475, + "step": 24570 + }, + { + "epoch": 0.758688807951108, + "grad_norm": 0.36636537313461304, + "learning_rate": 6.8613586764818e-05, + "loss": 0.0453, + "step": 24580 + }, + { + "epoch": 0.7589974689795667, + "grad_norm": 0.3687070310115814, + "learning_rate": 6.858800686127233e-05, + "loss": 0.0478, + "step": 24590 + }, + { + "epoch": 0.7593061300080252, + "grad_norm": 0.39759567379951477, + "learning_rate": 6.856242131118097e-05, + "loss": 0.0464, + "step": 24600 + }, + { + "epoch": 0.7596147910364838, + "grad_norm": 0.27891120314598083, + "learning_rate": 6.853683012231614e-05, + "loss": 0.0455, + "step": 24610 + }, + { + "epoch": 0.7599234520649423, + "grad_norm": 0.32833948731422424, + "learning_rate": 6.851123330245173e-05, + "loss": 0.0479, + "step": 24620 + }, + { + "epoch": 0.7602321130934008, + "grad_norm": 0.27108681201934814, + "learning_rate": 6.848563085936343e-05, + "loss": 0.0457, + "step": 24630 + }, + { + "epoch": 0.7605407741218594, + "grad_norm": 0.4024224877357483, + "learning_rate": 6.846002280082853e-05, + "loss": 0.0434, + "step": 24640 + }, + { + "epoch": 0.7608494351503179, + "grad_norm": 0.36702343821525574, + "learning_rate": 6.843440913462614e-05, + "loss": 0.0438, + "step": 24650 + }, + { + "epoch": 0.7611580961787765, + "grad_norm": 0.5048965215682983, + "learning_rate": 6.840878986853698e-05, + "loss": 0.0477, + "step": 24660 + }, + { + "epoch": 0.761466757207235, + "grad_norm": 0.312373548746109, + "learning_rate": 6.838316501034352e-05, + "loss": 0.041, + "step": 24670 + }, + { + "epoch": 0.7617754182356936, + "grad_norm": 0.2476939707994461, + "learning_rate": 6.83575345678299e-05, + "loss": 0.0431, + "step": 24680 + }, + { + "epoch": 0.7620840792641521, + "grad_norm": 0.3714669942855835, + "learning_rate": 6.833189854878196e-05, + "loss": 0.0484, + "step": 24690 + }, + { + "epoch": 0.7623927402926106, + "grad_norm": 0.38549813628196716, + "learning_rate": 6.83062569609873e-05, + "loss": 0.0456, + "step": 24700 + }, + { + "epoch": 0.7627014013210692, + "grad_norm": 0.6643197536468506, + "learning_rate": 6.828060981223512e-05, + "loss": 0.0476, + "step": 24710 + }, + { + "epoch": 0.7630100623495277, + "grad_norm": 0.3136754631996155, + "learning_rate": 6.825495711031634e-05, + "loss": 0.0401, + "step": 24720 + }, + { + "epoch": 0.7633187233779863, + "grad_norm": 0.322319895029068, + "learning_rate": 6.822929886302359e-05, + "loss": 0.0449, + "step": 24730 + }, + { + "epoch": 0.7636273844064448, + "grad_norm": 0.38875579833984375, + "learning_rate": 6.820363507815116e-05, + "loss": 0.048, + "step": 24740 + }, + { + "epoch": 0.7639360454349033, + "grad_norm": 0.34515196084976196, + "learning_rate": 6.817796576349501e-05, + "loss": 0.0461, + "step": 24750 + }, + { + "epoch": 0.764244706463362, + "grad_norm": 0.3519713282585144, + "learning_rate": 6.815229092685285e-05, + "loss": 0.0457, + "step": 24760 + }, + { + "epoch": 0.7645533674918205, + "grad_norm": 0.284647673368454, + "learning_rate": 6.812661057602399e-05, + "loss": 0.0478, + "step": 24770 + }, + { + "epoch": 0.7648620285202791, + "grad_norm": 0.3378578722476959, + "learning_rate": 6.810092471880943e-05, + "loss": 0.0431, + "step": 24780 + }, + { + "epoch": 0.7651706895487376, + "grad_norm": 0.332431823015213, + "learning_rate": 6.807523336301187e-05, + "loss": 0.046, + "step": 24790 + }, + { + "epoch": 0.7654793505771961, + "grad_norm": 0.2686578929424286, + "learning_rate": 6.804953651643566e-05, + "loss": 0.0487, + "step": 24800 + }, + { + "epoch": 0.7657880116056547, + "grad_norm": 0.26632583141326904, + "learning_rate": 6.802383418688685e-05, + "loss": 0.0453, + "step": 24810 + }, + { + "epoch": 0.7660966726341132, + "grad_norm": 0.379316121339798, + "learning_rate": 6.799812638217309e-05, + "loss": 0.0498, + "step": 24820 + }, + { + "epoch": 0.7664053336625718, + "grad_norm": 1.3151644468307495, + "learning_rate": 6.797241311010373e-05, + "loss": 0.0479, + "step": 24830 + }, + { + "epoch": 0.7667139946910303, + "grad_norm": 0.3312283158302307, + "learning_rate": 6.794669437848982e-05, + "loss": 0.0457, + "step": 24840 + }, + { + "epoch": 0.7670226557194889, + "grad_norm": 0.27050501108169556, + "learning_rate": 6.792097019514402e-05, + "loss": 0.0447, + "step": 24850 + }, + { + "epoch": 0.7673313167479474, + "grad_norm": 0.3714115023612976, + "learning_rate": 6.789524056788064e-05, + "loss": 0.0565, + "step": 24860 + }, + { + "epoch": 0.7676399777764059, + "grad_norm": 0.38429316878318787, + "learning_rate": 6.786950550451567e-05, + "loss": 0.0441, + "step": 24870 + }, + { + "epoch": 0.7679486388048645, + "grad_norm": 0.4159427881240845, + "learning_rate": 6.784376501286676e-05, + "loss": 0.0455, + "step": 24880 + }, + { + "epoch": 0.768257299833323, + "grad_norm": 0.40332403779029846, + "learning_rate": 6.781801910075316e-05, + "loss": 0.048, + "step": 24890 + }, + { + "epoch": 0.7685659608617816, + "grad_norm": 0.3671952486038208, + "learning_rate": 6.779226777599581e-05, + "loss": 0.0431, + "step": 24900 + }, + { + "epoch": 0.7688746218902401, + "grad_norm": 0.9161316156387329, + "learning_rate": 6.776651104641729e-05, + "loss": 0.0503, + "step": 24910 + }, + { + "epoch": 0.7691832829186986, + "grad_norm": 0.3673918843269348, + "learning_rate": 6.774074891984183e-05, + "loss": 0.0454, + "step": 24920 + }, + { + "epoch": 0.7694919439471573, + "grad_norm": 0.40751099586486816, + "learning_rate": 6.771498140409526e-05, + "loss": 0.0469, + "step": 24930 + }, + { + "epoch": 0.7698006049756158, + "grad_norm": 0.387271910905838, + "learning_rate": 6.768920850700506e-05, + "loss": 0.048, + "step": 24940 + }, + { + "epoch": 0.7701092660040744, + "grad_norm": 0.404900461435318, + "learning_rate": 6.766343023640039e-05, + "loss": 0.0439, + "step": 24950 + }, + { + "epoch": 0.7704179270325329, + "grad_norm": 0.380862295627594, + "learning_rate": 6.763764660011198e-05, + "loss": 0.0523, + "step": 24960 + }, + { + "epoch": 0.7707265880609914, + "grad_norm": 0.35366275906562805, + "learning_rate": 6.761185760597223e-05, + "loss": 0.046, + "step": 24970 + }, + { + "epoch": 0.77103524908945, + "grad_norm": 0.30033430457115173, + "learning_rate": 6.758606326181515e-05, + "loss": 0.0467, + "step": 24980 + }, + { + "epoch": 0.7713439101179085, + "grad_norm": 0.4254292845726013, + "learning_rate": 6.75602635754764e-05, + "loss": 0.0431, + "step": 24990 + }, + { + "epoch": 0.7716525711463671, + "grad_norm": 0.3398301899433136, + "learning_rate": 6.75344585547932e-05, + "loss": 0.0514, + "step": 25000 + }, + { + "epoch": 0.7719612321748256, + "grad_norm": 0.3517182469367981, + "learning_rate": 6.750864820760449e-05, + "loss": 0.0437, + "step": 25010 + }, + { + "epoch": 0.7722698932032842, + "grad_norm": 0.2769474387168884, + "learning_rate": 6.748283254175072e-05, + "loss": 0.0461, + "step": 25020 + }, + { + "epoch": 0.7725785542317427, + "grad_norm": 0.3411705493927002, + "learning_rate": 6.745701156507404e-05, + "loss": 0.0456, + "step": 25030 + }, + { + "epoch": 0.7728872152602012, + "grad_norm": 0.30775880813598633, + "learning_rate": 6.743118528541818e-05, + "loss": 0.0489, + "step": 25040 + }, + { + "epoch": 0.7731958762886598, + "grad_norm": 0.3509834408760071, + "learning_rate": 6.740535371062846e-05, + "loss": 0.0434, + "step": 25050 + }, + { + "epoch": 0.7735045373171183, + "grad_norm": 0.3607245981693268, + "learning_rate": 6.737951684855185e-05, + "loss": 0.0506, + "step": 25060 + }, + { + "epoch": 0.7738131983455769, + "grad_norm": 0.37024450302124023, + "learning_rate": 6.735367470703691e-05, + "loss": 0.0423, + "step": 25070 + }, + { + "epoch": 0.7741218593740354, + "grad_norm": 0.33287233114242554, + "learning_rate": 6.732782729393379e-05, + "loss": 0.0484, + "step": 25080 + }, + { + "epoch": 0.7744305204024939, + "grad_norm": 0.3933013677597046, + "learning_rate": 6.730197461709425e-05, + "loss": 0.0475, + "step": 25090 + }, + { + "epoch": 0.7747391814309526, + "grad_norm": 0.400488942861557, + "learning_rate": 6.727611668437164e-05, + "loss": 0.0442, + "step": 25100 + }, + { + "epoch": 0.775047842459411, + "grad_norm": 0.46040645241737366, + "learning_rate": 6.725025350362094e-05, + "loss": 0.0482, + "step": 25110 + }, + { + "epoch": 0.7753565034878697, + "grad_norm": 0.3848317563533783, + "learning_rate": 6.72243850826987e-05, + "loss": 0.045, + "step": 25120 + }, + { + "epoch": 0.7756651645163282, + "grad_norm": 0.3331156373023987, + "learning_rate": 6.719851142946305e-05, + "loss": 0.0468, + "step": 25130 + }, + { + "epoch": 0.7759738255447867, + "grad_norm": 0.4339883029460907, + "learning_rate": 6.717263255177372e-05, + "loss": 0.0481, + "step": 25140 + }, + { + "epoch": 0.7762824865732453, + "grad_norm": 0.2979801893234253, + "learning_rate": 6.714674845749205e-05, + "loss": 0.046, + "step": 25150 + }, + { + "epoch": 0.7765911476017038, + "grad_norm": 0.3348867893218994, + "learning_rate": 6.712085915448092e-05, + "loss": 0.0472, + "step": 25160 + }, + { + "epoch": 0.7768998086301624, + "grad_norm": 0.35013288259506226, + "learning_rate": 6.709496465060486e-05, + "loss": 0.0456, + "step": 25170 + }, + { + "epoch": 0.7772084696586209, + "grad_norm": 0.31704258918762207, + "learning_rate": 6.706906495372987e-05, + "loss": 0.0478, + "step": 25180 + }, + { + "epoch": 0.7775171306870794, + "grad_norm": 0.45092323422431946, + "learning_rate": 6.704316007172365e-05, + "loss": 0.0479, + "step": 25190 + }, + { + "epoch": 0.777825791715538, + "grad_norm": 0.3756072521209717, + "learning_rate": 6.701725001245539e-05, + "loss": 0.0444, + "step": 25200 + }, + { + "epoch": 0.7781344527439965, + "grad_norm": 0.385303795337677, + "learning_rate": 6.699133478379588e-05, + "loss": 0.0451, + "step": 25210 + }, + { + "epoch": 0.7784431137724551, + "grad_norm": 0.3938935697078705, + "learning_rate": 6.69654143936175e-05, + "loss": 0.0465, + "step": 25220 + }, + { + "epoch": 0.7787517748009136, + "grad_norm": 0.3842642307281494, + "learning_rate": 6.693948884979419e-05, + "loss": 0.0461, + "step": 25230 + }, + { + "epoch": 0.7790604358293722, + "grad_norm": 0.2844204604625702, + "learning_rate": 6.691355816020142e-05, + "loss": 0.0426, + "step": 25240 + }, + { + "epoch": 0.7793690968578307, + "grad_norm": 0.25327068567276, + "learning_rate": 6.688762233271624e-05, + "loss": 0.0484, + "step": 25250 + }, + { + "epoch": 0.7796777578862892, + "grad_norm": 0.44585487246513367, + "learning_rate": 6.68616813752173e-05, + "loss": 0.0463, + "step": 25260 + }, + { + "epoch": 0.7799864189147478, + "grad_norm": 0.2838420271873474, + "learning_rate": 6.683573529558477e-05, + "loss": 0.0472, + "step": 25270 + }, + { + "epoch": 0.7802950799432063, + "grad_norm": 0.46617141366004944, + "learning_rate": 6.680978410170037e-05, + "loss": 0.0409, + "step": 25280 + }, + { + "epoch": 0.780603740971665, + "grad_norm": 0.29378193616867065, + "learning_rate": 6.678382780144741e-05, + "loss": 0.0469, + "step": 25290 + }, + { + "epoch": 0.7809124020001235, + "grad_norm": 0.4771692752838135, + "learning_rate": 6.675786640271071e-05, + "loss": 0.043, + "step": 25300 + }, + { + "epoch": 0.781221063028582, + "grad_norm": 0.37509822845458984, + "learning_rate": 6.673189991337665e-05, + "loss": 0.0453, + "step": 25310 + }, + { + "epoch": 0.7815297240570406, + "grad_norm": 0.415998637676239, + "learning_rate": 6.670592834133317e-05, + "loss": 0.0446, + "step": 25320 + }, + { + "epoch": 0.7818383850854991, + "grad_norm": 0.35218682885169983, + "learning_rate": 6.667995169446979e-05, + "loss": 0.0471, + "step": 25330 + }, + { + "epoch": 0.7821470461139577, + "grad_norm": 0.3053489625453949, + "learning_rate": 6.665396998067747e-05, + "loss": 0.0431, + "step": 25340 + }, + { + "epoch": 0.7824557071424162, + "grad_norm": 0.3117581903934479, + "learning_rate": 6.66279832078488e-05, + "loss": 0.0483, + "step": 25350 + }, + { + "epoch": 0.7827643681708747, + "grad_norm": 0.3467715084552765, + "learning_rate": 6.660199138387786e-05, + "loss": 0.0452, + "step": 25360 + }, + { + "epoch": 0.7830730291993333, + "grad_norm": 0.3852631449699402, + "learning_rate": 6.65759945166603e-05, + "loss": 0.0473, + "step": 25370 + }, + { + "epoch": 0.7833816902277918, + "grad_norm": 0.5885218977928162, + "learning_rate": 6.654999261409326e-05, + "loss": 0.0459, + "step": 25380 + }, + { + "epoch": 0.7836903512562504, + "grad_norm": 0.33886849880218506, + "learning_rate": 6.652398568407544e-05, + "loss": 0.0463, + "step": 25390 + }, + { + "epoch": 0.7839990122847089, + "grad_norm": 0.3318488895893097, + "learning_rate": 6.649797373450707e-05, + "loss": 0.0435, + "step": 25400 + }, + { + "epoch": 0.7843076733131675, + "grad_norm": 0.39301472902297974, + "learning_rate": 6.647195677328988e-05, + "loss": 0.0492, + "step": 25410 + }, + { + "epoch": 0.784616334341626, + "grad_norm": 0.38151049613952637, + "learning_rate": 6.644593480832712e-05, + "loss": 0.0485, + "step": 25420 + }, + { + "epoch": 0.7849249953700845, + "grad_norm": 0.31984075903892517, + "learning_rate": 6.641990784752363e-05, + "loss": 0.0456, + "step": 25430 + }, + { + "epoch": 0.7852336563985431, + "grad_norm": 0.36825382709503174, + "learning_rate": 6.639387589878566e-05, + "loss": 0.0439, + "step": 25440 + }, + { + "epoch": 0.7855423174270016, + "grad_norm": 0.31528788805007935, + "learning_rate": 6.636783897002103e-05, + "loss": 0.051, + "step": 25450 + }, + { + "epoch": 0.7858509784554603, + "grad_norm": 0.3826563060283661, + "learning_rate": 6.63417970691391e-05, + "loss": 0.0429, + "step": 25460 + }, + { + "epoch": 0.7861596394839188, + "grad_norm": 0.3447008430957794, + "learning_rate": 6.63157502040507e-05, + "loss": 0.0486, + "step": 25470 + }, + { + "epoch": 0.7864683005123773, + "grad_norm": 0.4251173138618469, + "learning_rate": 6.628969838266819e-05, + "loss": 0.0526, + "step": 25480 + }, + { + "epoch": 0.7867769615408359, + "grad_norm": 0.371003657579422, + "learning_rate": 6.626364161290541e-05, + "loss": 0.0472, + "step": 25490 + }, + { + "epoch": 0.7870856225692944, + "grad_norm": 0.45678743720054626, + "learning_rate": 6.623757990267774e-05, + "loss": 0.0514, + "step": 25500 + }, + { + "epoch": 0.787394283597753, + "grad_norm": 0.35072028636932373, + "learning_rate": 6.621151325990201e-05, + "loss": 0.0451, + "step": 25510 + }, + { + "epoch": 0.7877029446262115, + "grad_norm": 0.45489585399627686, + "learning_rate": 6.618544169249657e-05, + "loss": 0.0479, + "step": 25520 + }, + { + "epoch": 0.78801160565467, + "grad_norm": 0.39243409037590027, + "learning_rate": 6.615936520838133e-05, + "loss": 0.0469, + "step": 25530 + }, + { + "epoch": 0.7883202666831286, + "grad_norm": 0.3623407483100891, + "learning_rate": 6.613328381547759e-05, + "loss": 0.0469, + "step": 25540 + }, + { + "epoch": 0.7886289277115871, + "grad_norm": 0.5466829538345337, + "learning_rate": 6.610719752170821e-05, + "loss": 0.0469, + "step": 25550 + }, + { + "epoch": 0.7889375887400457, + "grad_norm": 0.31359758973121643, + "learning_rate": 6.60811063349975e-05, + "loss": 0.0457, + "step": 25560 + }, + { + "epoch": 0.7892462497685042, + "grad_norm": 0.32875508069992065, + "learning_rate": 6.605501026327127e-05, + "loss": 0.0465, + "step": 25570 + }, + { + "epoch": 0.7895549107969628, + "grad_norm": 0.37886205315589905, + "learning_rate": 6.602890931445685e-05, + "loss": 0.0461, + "step": 25580 + }, + { + "epoch": 0.7898635718254213, + "grad_norm": 0.36717620491981506, + "learning_rate": 6.6002803496483e-05, + "loss": 0.0522, + "step": 25590 + }, + { + "epoch": 0.7901722328538798, + "grad_norm": 0.31724873185157776, + "learning_rate": 6.597669281727997e-05, + "loss": 0.048, + "step": 25600 + }, + { + "epoch": 0.7904808938823384, + "grad_norm": 0.3093792200088501, + "learning_rate": 6.595057728477949e-05, + "loss": 0.042, + "step": 25610 + }, + { + "epoch": 0.7907895549107969, + "grad_norm": 0.27069732546806335, + "learning_rate": 6.59244569069148e-05, + "loss": 0.0453, + "step": 25620 + }, + { + "epoch": 0.7910982159392556, + "grad_norm": 0.5177167057991028, + "learning_rate": 6.589833169162054e-05, + "loss": 0.0455, + "step": 25630 + }, + { + "epoch": 0.7914068769677141, + "grad_norm": 0.35792276263237, + "learning_rate": 6.587220164683291e-05, + "loss": 0.0433, + "step": 25640 + }, + { + "epoch": 0.7917155379961726, + "grad_norm": 0.2945466637611389, + "learning_rate": 6.58460667804895e-05, + "loss": 0.0469, + "step": 25650 + }, + { + "epoch": 0.7920241990246312, + "grad_norm": 0.3765825927257538, + "learning_rate": 6.581992710052938e-05, + "loss": 0.0456, + "step": 25660 + }, + { + "epoch": 0.7923328600530897, + "grad_norm": 0.3265666365623474, + "learning_rate": 6.579378261489311e-05, + "loss": 0.0475, + "step": 25670 + }, + { + "epoch": 0.7926415210815483, + "grad_norm": 0.3505082130432129, + "learning_rate": 6.576763333152268e-05, + "loss": 0.0463, + "step": 25680 + }, + { + "epoch": 0.7929501821100068, + "grad_norm": 0.36164215207099915, + "learning_rate": 6.574147925836159e-05, + "loss": 0.0496, + "step": 25690 + }, + { + "epoch": 0.7932588431384653, + "grad_norm": 0.3248909115791321, + "learning_rate": 6.571532040335472e-05, + "loss": 0.0477, + "step": 25700 + }, + { + "epoch": 0.7935675041669239, + "grad_norm": 0.4262208640575409, + "learning_rate": 6.568915677444845e-05, + "loss": 0.0496, + "step": 25710 + }, + { + "epoch": 0.7938761651953824, + "grad_norm": 0.3198816478252411, + "learning_rate": 6.56629883795906e-05, + "loss": 0.0466, + "step": 25720 + }, + { + "epoch": 0.794184826223841, + "grad_norm": 0.60462486743927, + "learning_rate": 6.563681522673043e-05, + "loss": 0.044, + "step": 25730 + }, + { + "epoch": 0.7944934872522995, + "grad_norm": 0.268476277589798, + "learning_rate": 6.561063732381867e-05, + "loss": 0.0476, + "step": 25740 + }, + { + "epoch": 0.7948021482807581, + "grad_norm": 0.4646087884902954, + "learning_rate": 6.558445467880745e-05, + "loss": 0.0444, + "step": 25750 + }, + { + "epoch": 0.7951108093092166, + "grad_norm": 0.5858204364776611, + "learning_rate": 6.55582672996504e-05, + "loss": 0.0445, + "step": 25760 + }, + { + "epoch": 0.7954194703376751, + "grad_norm": 0.34128710627555847, + "learning_rate": 6.553207519430253e-05, + "loss": 0.0466, + "step": 25770 + }, + { + "epoch": 0.7957281313661337, + "grad_norm": 0.3079786002635956, + "learning_rate": 6.550587837072032e-05, + "loss": 0.0458, + "step": 25780 + }, + { + "epoch": 0.7960367923945922, + "grad_norm": 0.3940736651420593, + "learning_rate": 6.547967683686166e-05, + "loss": 0.054, + "step": 25790 + }, + { + "epoch": 0.7963454534230509, + "grad_norm": 0.3133600950241089, + "learning_rate": 6.545347060068591e-05, + "loss": 0.0443, + "step": 25800 + }, + { + "epoch": 0.7966541144515094, + "grad_norm": 0.4558347761631012, + "learning_rate": 6.542725967015382e-05, + "loss": 0.0456, + "step": 25810 + }, + { + "epoch": 0.7969627754799679, + "grad_norm": 0.3490212857723236, + "learning_rate": 6.540104405322757e-05, + "loss": 0.0459, + "step": 25820 + }, + { + "epoch": 0.7972714365084265, + "grad_norm": 0.39668455719947815, + "learning_rate": 6.537482375787077e-05, + "loss": 0.0455, + "step": 25830 + }, + { + "epoch": 0.797580097536885, + "grad_norm": 0.38529324531555176, + "learning_rate": 6.534859879204845e-05, + "loss": 0.0496, + "step": 25840 + }, + { + "epoch": 0.7978887585653436, + "grad_norm": 0.2712022066116333, + "learning_rate": 6.532236916372709e-05, + "loss": 0.0433, + "step": 25850 + }, + { + "epoch": 0.7981974195938021, + "grad_norm": 0.371285080909729, + "learning_rate": 6.529613488087454e-05, + "loss": 0.047, + "step": 25860 + }, + { + "epoch": 0.7985060806222606, + "grad_norm": 0.3944352865219116, + "learning_rate": 6.526989595146009e-05, + "loss": 0.0485, + "step": 25870 + }, + { + "epoch": 0.7988147416507192, + "grad_norm": 0.37070608139038086, + "learning_rate": 6.524365238345441e-05, + "loss": 0.045, + "step": 25880 + }, + { + "epoch": 0.7991234026791777, + "grad_norm": 0.40851110219955444, + "learning_rate": 6.521740418482964e-05, + "loss": 0.049, + "step": 25890 + }, + { + "epoch": 0.7994320637076363, + "grad_norm": 0.30000039935112, + "learning_rate": 6.519115136355925e-05, + "loss": 0.0477, + "step": 25900 + }, + { + "epoch": 0.7997407247360948, + "grad_norm": 0.38303765654563904, + "learning_rate": 6.51648939276182e-05, + "loss": 0.0417, + "step": 25910 + }, + { + "epoch": 0.8000493857645534, + "grad_norm": 0.3213516175746918, + "learning_rate": 6.513863188498277e-05, + "loss": 0.0437, + "step": 25920 + }, + { + "epoch": 0.8003580467930119, + "grad_norm": 0.3223700225353241, + "learning_rate": 6.511236524363068e-05, + "loss": 0.0472, + "step": 25930 + }, + { + "epoch": 0.8006667078214704, + "grad_norm": 0.44971561431884766, + "learning_rate": 6.508609401154104e-05, + "loss": 0.0437, + "step": 25940 + }, + { + "epoch": 0.800975368849929, + "grad_norm": 0.42918387055397034, + "learning_rate": 6.505981819669439e-05, + "loss": 0.0432, + "step": 25950 + }, + { + "epoch": 0.8012840298783875, + "grad_norm": 0.2873128354549408, + "learning_rate": 6.503353780707258e-05, + "loss": 0.046, + "step": 25960 + }, + { + "epoch": 0.8015926909068462, + "grad_norm": 0.27158087491989136, + "learning_rate": 6.500725285065895e-05, + "loss": 0.0434, + "step": 25970 + }, + { + "epoch": 0.8019013519353047, + "grad_norm": 0.37709298729896545, + "learning_rate": 6.498096333543813e-05, + "loss": 0.0493, + "step": 25980 + }, + { + "epoch": 0.8022100129637632, + "grad_norm": 0.29693445563316345, + "learning_rate": 6.49546692693962e-05, + "loss": 0.046, + "step": 25990 + }, + { + "epoch": 0.8025186739922218, + "grad_norm": 0.4073707163333893, + "learning_rate": 6.492837066052059e-05, + "loss": 0.0437, + "step": 26000 + }, + { + "epoch": 0.8028273350206803, + "grad_norm": 0.35951346158981323, + "learning_rate": 6.490206751680014e-05, + "loss": 0.0478, + "step": 26010 + }, + { + "epoch": 0.8031359960491389, + "grad_norm": 0.39018696546554565, + "learning_rate": 6.487575984622505e-05, + "loss": 0.0446, + "step": 26020 + }, + { + "epoch": 0.8034446570775974, + "grad_norm": 0.37592557072639465, + "learning_rate": 6.484944765678689e-05, + "loss": 0.0454, + "step": 26030 + }, + { + "epoch": 0.8037533181060559, + "grad_norm": 0.28809210658073425, + "learning_rate": 6.482313095647861e-05, + "loss": 0.0461, + "step": 26040 + }, + { + "epoch": 0.8040619791345145, + "grad_norm": 0.2891172170639038, + "learning_rate": 6.479680975329451e-05, + "loss": 0.0437, + "step": 26050 + }, + { + "epoch": 0.804370640162973, + "grad_norm": 0.33722418546676636, + "learning_rate": 6.477048405523031e-05, + "loss": 0.052, + "step": 26060 + }, + { + "epoch": 0.8046793011914316, + "grad_norm": 0.38875991106033325, + "learning_rate": 6.474415387028304e-05, + "loss": 0.0474, + "step": 26070 + }, + { + "epoch": 0.8049879622198901, + "grad_norm": 0.3684637248516083, + "learning_rate": 6.471781920645114e-05, + "loss": 0.0465, + "step": 26080 + }, + { + "epoch": 0.8052966232483486, + "grad_norm": 0.34935662150382996, + "learning_rate": 6.469148007173434e-05, + "loss": 0.0419, + "step": 26090 + }, + { + "epoch": 0.8056052842768072, + "grad_norm": 0.48826420307159424, + "learning_rate": 6.466513647413381e-05, + "loss": 0.0451, + "step": 26100 + }, + { + "epoch": 0.8059139453052657, + "grad_norm": 0.6680077314376831, + "learning_rate": 6.463878842165203e-05, + "loss": 0.0499, + "step": 26110 + }, + { + "epoch": 0.8062226063337243, + "grad_norm": 0.31191545724868774, + "learning_rate": 6.461243592229286e-05, + "loss": 0.0497, + "step": 26120 + }, + { + "epoch": 0.8065312673621828, + "grad_norm": 0.40696612000465393, + "learning_rate": 6.458607898406146e-05, + "loss": 0.0457, + "step": 26130 + }, + { + "epoch": 0.8068399283906414, + "grad_norm": 0.35318809747695923, + "learning_rate": 6.455971761496439e-05, + "loss": 0.044, + "step": 26140 + }, + { + "epoch": 0.8071485894191, + "grad_norm": 0.7270426154136658, + "learning_rate": 6.453335182300953e-05, + "loss": 0.0434, + "step": 26150 + }, + { + "epoch": 0.8074572504475585, + "grad_norm": 0.3302808701992035, + "learning_rate": 6.450698161620612e-05, + "loss": 0.0445, + "step": 26160 + }, + { + "epoch": 0.8077659114760171, + "grad_norm": 0.33275917172431946, + "learning_rate": 6.448060700256473e-05, + "loss": 0.048, + "step": 26170 + }, + { + "epoch": 0.8080745725044756, + "grad_norm": 0.2698535919189453, + "learning_rate": 6.445422799009726e-05, + "loss": 0.0419, + "step": 26180 + }, + { + "epoch": 0.8083832335329342, + "grad_norm": 0.2677537202835083, + "learning_rate": 6.442784458681699e-05, + "loss": 0.0458, + "step": 26190 + }, + { + "epoch": 0.8086918945613927, + "grad_norm": 0.3169616162776947, + "learning_rate": 6.440145680073847e-05, + "loss": 0.046, + "step": 26200 + }, + { + "epoch": 0.8090005555898512, + "grad_norm": 0.2949278652667999, + "learning_rate": 6.437506463987762e-05, + "loss": 0.0453, + "step": 26210 + }, + { + "epoch": 0.8093092166183098, + "grad_norm": 0.33292457461357117, + "learning_rate": 6.434866811225168e-05, + "loss": 0.049, + "step": 26220 + }, + { + "epoch": 0.8096178776467683, + "grad_norm": 0.2976985573768616, + "learning_rate": 6.432226722587923e-05, + "loss": 0.0441, + "step": 26230 + }, + { + "epoch": 0.8099265386752269, + "grad_norm": 0.34964045882225037, + "learning_rate": 6.429586198878015e-05, + "loss": 0.0484, + "step": 26240 + }, + { + "epoch": 0.8102351997036854, + "grad_norm": 0.3654073178768158, + "learning_rate": 6.426945240897566e-05, + "loss": 0.0411, + "step": 26250 + }, + { + "epoch": 0.8105438607321439, + "grad_norm": 0.3142799139022827, + "learning_rate": 6.424303849448829e-05, + "loss": 0.0491, + "step": 26260 + }, + { + "epoch": 0.8108525217606025, + "grad_norm": 0.34440183639526367, + "learning_rate": 6.42166202533419e-05, + "loss": 0.0465, + "step": 26270 + }, + { + "epoch": 0.811161182789061, + "grad_norm": 0.3443383276462555, + "learning_rate": 6.419019769356164e-05, + "loss": 0.0476, + "step": 26280 + }, + { + "epoch": 0.8114698438175196, + "grad_norm": 0.3426738381385803, + "learning_rate": 6.416377082317398e-05, + "loss": 0.0444, + "step": 26290 + }, + { + "epoch": 0.8117785048459781, + "grad_norm": 0.4231814742088318, + "learning_rate": 6.413733965020674e-05, + "loss": 0.0398, + "step": 26300 + }, + { + "epoch": 0.8120871658744367, + "grad_norm": 0.46044817566871643, + "learning_rate": 6.411090418268896e-05, + "loss": 0.0502, + "step": 26310 + }, + { + "epoch": 0.8123958269028952, + "grad_norm": 0.3165290653705597, + "learning_rate": 6.408446442865109e-05, + "loss": 0.0455, + "step": 26320 + }, + { + "epoch": 0.8127044879313537, + "grad_norm": 0.5166967511177063, + "learning_rate": 6.405802039612479e-05, + "loss": 0.0479, + "step": 26330 + }, + { + "epoch": 0.8130131489598124, + "grad_norm": 0.3266200125217438, + "learning_rate": 6.403157209314308e-05, + "loss": 0.0518, + "step": 26340 + }, + { + "epoch": 0.8133218099882709, + "grad_norm": 0.34235692024230957, + "learning_rate": 6.400511952774024e-05, + "loss": 0.0458, + "step": 26350 + }, + { + "epoch": 0.8136304710167295, + "grad_norm": 0.36672258377075195, + "learning_rate": 6.397866270795187e-05, + "loss": 0.0424, + "step": 26360 + }, + { + "epoch": 0.813939132045188, + "grad_norm": 0.35113024711608887, + "learning_rate": 6.395220164181489e-05, + "loss": 0.0487, + "step": 26370 + }, + { + "epoch": 0.8142477930736465, + "grad_norm": 0.3485586643218994, + "learning_rate": 6.39257363373674e-05, + "loss": 0.0493, + "step": 26380 + }, + { + "epoch": 0.8145564541021051, + "grad_norm": 0.32418927550315857, + "learning_rate": 6.389926680264892e-05, + "loss": 0.05, + "step": 26390 + }, + { + "epoch": 0.8148651151305636, + "grad_norm": 0.39064985513687134, + "learning_rate": 6.387279304570017e-05, + "loss": 0.049, + "step": 26400 + }, + { + "epoch": 0.8151737761590222, + "grad_norm": 0.2911173403263092, + "learning_rate": 6.384631507456319e-05, + "loss": 0.0447, + "step": 26410 + }, + { + "epoch": 0.8154824371874807, + "grad_norm": 0.382009357213974, + "learning_rate": 6.381983289728126e-05, + "loss": 0.0439, + "step": 26420 + }, + { + "epoch": 0.8157910982159392, + "grad_norm": 0.5013092756271362, + "learning_rate": 6.3793346521899e-05, + "loss": 0.0448, + "step": 26430 + }, + { + "epoch": 0.8160997592443978, + "grad_norm": 0.29794248938560486, + "learning_rate": 6.376685595646226e-05, + "loss": 0.0455, + "step": 26440 + }, + { + "epoch": 0.8164084202728563, + "grad_norm": 0.3153747618198395, + "learning_rate": 6.374036120901816e-05, + "loss": 0.0418, + "step": 26450 + }, + { + "epoch": 0.8167170813013149, + "grad_norm": 0.41237255930900574, + "learning_rate": 6.371386228761514e-05, + "loss": 0.0402, + "step": 26460 + }, + { + "epoch": 0.8170257423297734, + "grad_norm": 0.32836073637008667, + "learning_rate": 6.368735920030283e-05, + "loss": 0.0459, + "step": 26470 + }, + { + "epoch": 0.817334403358232, + "grad_norm": 0.2866571843624115, + "learning_rate": 6.366085195513218e-05, + "loss": 0.0455, + "step": 26480 + }, + { + "epoch": 0.8176430643866905, + "grad_norm": 0.26351505517959595, + "learning_rate": 6.363434056015543e-05, + "loss": 0.0426, + "step": 26490 + }, + { + "epoch": 0.817951725415149, + "grad_norm": 0.361929714679718, + "learning_rate": 6.360782502342599e-05, + "loss": 0.0503, + "step": 26500 + }, + { + "epoch": 0.8182603864436077, + "grad_norm": 0.3754512071609497, + "learning_rate": 6.358130535299862e-05, + "loss": 0.0465, + "step": 26510 + }, + { + "epoch": 0.8185690474720662, + "grad_norm": 0.3740057945251465, + "learning_rate": 6.355478155692926e-05, + "loss": 0.0435, + "step": 26520 + }, + { + "epoch": 0.8188777085005248, + "grad_norm": 0.3456493616104126, + "learning_rate": 6.352825364327517e-05, + "loss": 0.0472, + "step": 26530 + }, + { + "epoch": 0.8191863695289833, + "grad_norm": 0.29335910081863403, + "learning_rate": 6.350172162009482e-05, + "loss": 0.0443, + "step": 26540 + }, + { + "epoch": 0.8194950305574418, + "grad_norm": 0.34719598293304443, + "learning_rate": 6.347518549544793e-05, + "loss": 0.0479, + "step": 26550 + }, + { + "epoch": 0.8198036915859004, + "grad_norm": 0.2886597514152527, + "learning_rate": 6.344864527739547e-05, + "loss": 0.0473, + "step": 26560 + }, + { + "epoch": 0.8201123526143589, + "grad_norm": 0.2986287772655487, + "learning_rate": 6.342210097399966e-05, + "loss": 0.0529, + "step": 26570 + }, + { + "epoch": 0.8204210136428175, + "grad_norm": 0.4054427146911621, + "learning_rate": 6.339555259332398e-05, + "loss": 0.0501, + "step": 26580 + }, + { + "epoch": 0.820729674671276, + "grad_norm": 0.501024603843689, + "learning_rate": 6.33690001434331e-05, + "loss": 0.0402, + "step": 26590 + }, + { + "epoch": 0.8210383356997345, + "grad_norm": 0.3504687547683716, + "learning_rate": 6.334244363239296e-05, + "loss": 0.0508, + "step": 26600 + }, + { + "epoch": 0.8213469967281931, + "grad_norm": 0.41193071007728577, + "learning_rate": 6.331588306827073e-05, + "loss": 0.0453, + "step": 26610 + }, + { + "epoch": 0.8216556577566516, + "grad_norm": 0.6084991693496704, + "learning_rate": 6.328931845913483e-05, + "loss": 0.0419, + "step": 26620 + }, + { + "epoch": 0.8219643187851102, + "grad_norm": 0.3968687355518341, + "learning_rate": 6.326274981305484e-05, + "loss": 0.0523, + "step": 26630 + }, + { + "epoch": 0.8222729798135687, + "grad_norm": 0.3521173894405365, + "learning_rate": 6.323617713810166e-05, + "loss": 0.0496, + "step": 26640 + }, + { + "epoch": 0.8225816408420273, + "grad_norm": 0.33273157477378845, + "learning_rate": 6.320960044234734e-05, + "loss": 0.0444, + "step": 26650 + }, + { + "epoch": 0.8228903018704858, + "grad_norm": 0.4159906506538391, + "learning_rate": 6.318301973386518e-05, + "loss": 0.0456, + "step": 26660 + }, + { + "epoch": 0.8231989628989443, + "grad_norm": 0.3391857445240021, + "learning_rate": 6.315643502072971e-05, + "loss": 0.046, + "step": 26670 + }, + { + "epoch": 0.823507623927403, + "grad_norm": 0.3922361433506012, + "learning_rate": 6.312984631101667e-05, + "loss": 0.0466, + "step": 26680 + }, + { + "epoch": 0.8238162849558615, + "grad_norm": 0.27028223872184753, + "learning_rate": 6.310325361280297e-05, + "loss": 0.0471, + "step": 26690 + }, + { + "epoch": 0.8241249459843201, + "grad_norm": 0.3524101674556732, + "learning_rate": 6.30766569341668e-05, + "loss": 0.0464, + "step": 26700 + }, + { + "epoch": 0.8244336070127786, + "grad_norm": 0.3584372103214264, + "learning_rate": 6.305005628318753e-05, + "loss": 0.044, + "step": 26710 + }, + { + "epoch": 0.8247422680412371, + "grad_norm": 0.3000434637069702, + "learning_rate": 6.302345166794572e-05, + "loss": 0.045, + "step": 26720 + }, + { + "epoch": 0.8250509290696957, + "grad_norm": 0.4004782736301422, + "learning_rate": 6.299684309652316e-05, + "loss": 0.0483, + "step": 26730 + }, + { + "epoch": 0.8253595900981542, + "grad_norm": 0.37783536314964294, + "learning_rate": 6.297023057700283e-05, + "loss": 0.0424, + "step": 26740 + }, + { + "epoch": 0.8256682511266128, + "grad_norm": 0.3428926169872284, + "learning_rate": 6.294361411746891e-05, + "loss": 0.0459, + "step": 26750 + }, + { + "epoch": 0.8259769121550713, + "grad_norm": 0.32275375723838806, + "learning_rate": 6.291699372600677e-05, + "loss": 0.0452, + "step": 26760 + }, + { + "epoch": 0.8262855731835298, + "grad_norm": 0.431611031293869, + "learning_rate": 6.2890369410703e-05, + "loss": 0.045, + "step": 26770 + }, + { + "epoch": 0.8265942342119884, + "grad_norm": 0.3628222644329071, + "learning_rate": 6.286374117964534e-05, + "loss": 0.0423, + "step": 26780 + }, + { + "epoch": 0.8269028952404469, + "grad_norm": 0.41734495759010315, + "learning_rate": 6.283710904092277e-05, + "loss": 0.0517, + "step": 26790 + }, + { + "epoch": 0.8272115562689055, + "grad_norm": 0.44290798902511597, + "learning_rate": 6.281047300262542e-05, + "loss": 0.0504, + "step": 26800 + }, + { + "epoch": 0.827520217297364, + "grad_norm": 0.2845461368560791, + "learning_rate": 6.278383307284461e-05, + "loss": 0.0432, + "step": 26810 + }, + { + "epoch": 0.8278288783258226, + "grad_norm": 0.33180680871009827, + "learning_rate": 6.275718925967284e-05, + "loss": 0.0441, + "step": 26820 + }, + { + "epoch": 0.8281375393542811, + "grad_norm": 0.3332653343677521, + "learning_rate": 6.273054157120382e-05, + "loss": 0.0467, + "step": 26830 + }, + { + "epoch": 0.8284462003827396, + "grad_norm": 0.3246449828147888, + "learning_rate": 6.270389001553238e-05, + "loss": 0.0407, + "step": 26840 + }, + { + "epoch": 0.8287548614111983, + "grad_norm": 0.3671886622905731, + "learning_rate": 6.26772346007546e-05, + "loss": 0.0463, + "step": 26850 + }, + { + "epoch": 0.8290635224396568, + "grad_norm": 0.33733007311820984, + "learning_rate": 6.265057533496767e-05, + "loss": 0.0488, + "step": 26860 + }, + { + "epoch": 0.8293721834681154, + "grad_norm": 0.3024577796459198, + "learning_rate": 6.262391222626997e-05, + "loss": 0.0427, + "step": 26870 + }, + { + "epoch": 0.8296808444965739, + "grad_norm": 0.36618852615356445, + "learning_rate": 6.259724528276106e-05, + "loss": 0.0441, + "step": 26880 + }, + { + "epoch": 0.8299895055250324, + "grad_norm": 0.40870732069015503, + "learning_rate": 6.257057451254162e-05, + "loss": 0.0443, + "step": 26890 + }, + { + "epoch": 0.830298166553491, + "grad_norm": 0.3730919659137726, + "learning_rate": 6.254389992371357e-05, + "loss": 0.0445, + "step": 26900 + }, + { + "epoch": 0.8306068275819495, + "grad_norm": 0.41742128133773804, + "learning_rate": 6.25172215243799e-05, + "loss": 0.0446, + "step": 26910 + }, + { + "epoch": 0.8309154886104081, + "grad_norm": 0.36159706115722656, + "learning_rate": 6.249053932264486e-05, + "loss": 0.0444, + "step": 26920 + }, + { + "epoch": 0.8312241496388666, + "grad_norm": 0.3049677312374115, + "learning_rate": 6.246385332661376e-05, + "loss": 0.0423, + "step": 26930 + }, + { + "epoch": 0.8315328106673251, + "grad_norm": 0.3503056466579437, + "learning_rate": 6.24371635443931e-05, + "loss": 0.0412, + "step": 26940 + }, + { + "epoch": 0.8318414716957837, + "grad_norm": 0.2696658968925476, + "learning_rate": 6.241046998409054e-05, + "loss": 0.0425, + "step": 26950 + }, + { + "epoch": 0.8321501327242422, + "grad_norm": 0.36879149079322815, + "learning_rate": 6.238377265381489e-05, + "loss": 0.0479, + "step": 26960 + }, + { + "epoch": 0.8324587937527008, + "grad_norm": 0.3504766523838043, + "learning_rate": 6.235707156167607e-05, + "loss": 0.0415, + "step": 26970 + }, + { + "epoch": 0.8327674547811593, + "grad_norm": 0.4345758259296417, + "learning_rate": 6.233036671578519e-05, + "loss": 0.0486, + "step": 26980 + }, + { + "epoch": 0.8330761158096178, + "grad_norm": 0.31077608466148376, + "learning_rate": 6.230365812425445e-05, + "loss": 0.0473, + "step": 26990 + }, + { + "epoch": 0.8333847768380764, + "grad_norm": 0.3557131290435791, + "learning_rate": 6.227694579519724e-05, + "loss": 0.047, + "step": 27000 + }, + { + "epoch": 0.8336934378665349, + "grad_norm": 0.37612584233283997, + "learning_rate": 6.225022973672805e-05, + "loss": 0.0448, + "step": 27010 + }, + { + "epoch": 0.8340020988949935, + "grad_norm": 0.36429232358932495, + "learning_rate": 6.222350995696253e-05, + "loss": 0.0437, + "step": 27020 + }, + { + "epoch": 0.834310759923452, + "grad_norm": 0.35749441385269165, + "learning_rate": 6.21967864640174e-05, + "loss": 0.0444, + "step": 27030 + }, + { + "epoch": 0.8346194209519107, + "grad_norm": 0.32837313413619995, + "learning_rate": 6.217005926601059e-05, + "loss": 0.0451, + "step": 27040 + }, + { + "epoch": 0.8349280819803692, + "grad_norm": 0.4702328145503998, + "learning_rate": 6.214332837106111e-05, + "loss": 0.0458, + "step": 27050 + }, + { + "epoch": 0.8352367430088277, + "grad_norm": 0.2816105782985687, + "learning_rate": 6.21165937872891e-05, + "loss": 0.0452, + "step": 27060 + }, + { + "epoch": 0.8355454040372863, + "grad_norm": 0.34524083137512207, + "learning_rate": 6.208985552281582e-05, + "loss": 0.0507, + "step": 27070 + }, + { + "epoch": 0.8358540650657448, + "grad_norm": 0.48269206285476685, + "learning_rate": 6.206311358576364e-05, + "loss": 0.0468, + "step": 27080 + }, + { + "epoch": 0.8361627260942034, + "grad_norm": 0.32092052698135376, + "learning_rate": 6.203636798425608e-05, + "loss": 0.0427, + "step": 27090 + }, + { + "epoch": 0.8364713871226619, + "grad_norm": 0.42769578099250793, + "learning_rate": 6.20096187264177e-05, + "loss": 0.0448, + "step": 27100 + }, + { + "epoch": 0.8367800481511204, + "grad_norm": 0.3634469211101532, + "learning_rate": 6.198286582037425e-05, + "loss": 0.0429, + "step": 27110 + }, + { + "epoch": 0.837088709179579, + "grad_norm": 0.31360456347465515, + "learning_rate": 6.195610927425256e-05, + "loss": 0.043, + "step": 27120 + }, + { + "epoch": 0.8373973702080375, + "grad_norm": 0.3736761808395386, + "learning_rate": 6.192934909618056e-05, + "loss": 0.0473, + "step": 27130 + }, + { + "epoch": 0.8377060312364961, + "grad_norm": 0.29345497488975525, + "learning_rate": 6.190258529428728e-05, + "loss": 0.0425, + "step": 27140 + }, + { + "epoch": 0.8380146922649546, + "grad_norm": 0.30796557664871216, + "learning_rate": 6.187581787670285e-05, + "loss": 0.0432, + "step": 27150 + }, + { + "epoch": 0.8383233532934131, + "grad_norm": 0.3877561688423157, + "learning_rate": 6.184904685155852e-05, + "loss": 0.04, + "step": 27160 + }, + { + "epoch": 0.8386320143218717, + "grad_norm": 0.3757551908493042, + "learning_rate": 6.18222722269866e-05, + "loss": 0.0447, + "step": 27170 + }, + { + "epoch": 0.8389406753503302, + "grad_norm": 0.3320651352405548, + "learning_rate": 6.179549401112053e-05, + "loss": 0.0457, + "step": 27180 + }, + { + "epoch": 0.8392493363787888, + "grad_norm": 0.3845139741897583, + "learning_rate": 6.176871221209482e-05, + "loss": 0.0457, + "step": 27190 + }, + { + "epoch": 0.8395579974072473, + "grad_norm": 0.263582706451416, + "learning_rate": 6.174192683804508e-05, + "loss": 0.0445, + "step": 27200 + }, + { + "epoch": 0.839866658435706, + "grad_norm": 0.25030261278152466, + "learning_rate": 6.1715137897108e-05, + "loss": 0.0456, + "step": 27210 + }, + { + "epoch": 0.8401753194641645, + "grad_norm": 0.803683876991272, + "learning_rate": 6.168834539742134e-05, + "loss": 0.0457, + "step": 27220 + }, + { + "epoch": 0.840483980492623, + "grad_norm": 0.35738545656204224, + "learning_rate": 6.166154934712397e-05, + "loss": 0.043, + "step": 27230 + }, + { + "epoch": 0.8407926415210816, + "grad_norm": 0.4948538839817047, + "learning_rate": 6.163474975435581e-05, + "loss": 0.0438, + "step": 27240 + }, + { + "epoch": 0.8411013025495401, + "grad_norm": 0.3864251673221588, + "learning_rate": 6.160794662725787e-05, + "loss": 0.046, + "step": 27250 + }, + { + "epoch": 0.8414099635779987, + "grad_norm": 0.28637829422950745, + "learning_rate": 6.158113997397222e-05, + "loss": 0.0442, + "step": 27260 + }, + { + "epoch": 0.8417186246064572, + "grad_norm": 0.3490058481693268, + "learning_rate": 6.155432980264205e-05, + "loss": 0.0433, + "step": 27270 + }, + { + "epoch": 0.8420272856349157, + "grad_norm": 0.26630252599716187, + "learning_rate": 6.152751612141156e-05, + "loss": 0.0473, + "step": 27280 + }, + { + "epoch": 0.8423359466633743, + "grad_norm": 0.35329654812812805, + "learning_rate": 6.150069893842602e-05, + "loss": 0.0441, + "step": 27290 + }, + { + "epoch": 0.8426446076918328, + "grad_norm": 0.37019556760787964, + "learning_rate": 6.147387826183182e-05, + "loss": 0.0461, + "step": 27300 + }, + { + "epoch": 0.8429532687202914, + "grad_norm": 0.3955848813056946, + "learning_rate": 6.144705409977635e-05, + "loss": 0.0448, + "step": 27310 + }, + { + "epoch": 0.8432619297487499, + "grad_norm": 0.2860700488090515, + "learning_rate": 6.142022646040808e-05, + "loss": 0.0467, + "step": 27320 + }, + { + "epoch": 0.8435705907772084, + "grad_norm": 0.31695330142974854, + "learning_rate": 6.139339535187653e-05, + "loss": 0.044, + "step": 27330 + }, + { + "epoch": 0.843879251805667, + "grad_norm": 0.32202309370040894, + "learning_rate": 6.136656078233232e-05, + "loss": 0.0479, + "step": 27340 + }, + { + "epoch": 0.8441879128341255, + "grad_norm": 0.36338427662849426, + "learning_rate": 6.133972275992707e-05, + "loss": 0.0449, + "step": 27350 + }, + { + "epoch": 0.8444965738625841, + "grad_norm": 0.42620766162872314, + "learning_rate": 6.131288129281342e-05, + "loss": 0.0403, + "step": 27360 + }, + { + "epoch": 0.8448052348910426, + "grad_norm": 0.3504875600337982, + "learning_rate": 6.128603638914516e-05, + "loss": 0.043, + "step": 27370 + }, + { + "epoch": 0.8451138959195013, + "grad_norm": 0.3122507631778717, + "learning_rate": 6.125918805707704e-05, + "loss": 0.0434, + "step": 27380 + }, + { + "epoch": 0.8454225569479598, + "grad_norm": 0.3409575819969177, + "learning_rate": 6.123233630476485e-05, + "loss": 0.0417, + "step": 27390 + }, + { + "epoch": 0.8457312179764183, + "grad_norm": 0.32262513041496277, + "learning_rate": 6.120548114036547e-05, + "loss": 0.0472, + "step": 27400 + }, + { + "epoch": 0.8460398790048769, + "grad_norm": 0.2836433947086334, + "learning_rate": 6.117862257203679e-05, + "loss": 0.0438, + "step": 27410 + }, + { + "epoch": 0.8463485400333354, + "grad_norm": 0.2869862914085388, + "learning_rate": 6.115176060793771e-05, + "loss": 0.041, + "step": 27420 + }, + { + "epoch": 0.846657201061794, + "grad_norm": 0.4259384870529175, + "learning_rate": 6.112489525622822e-05, + "loss": 0.0455, + "step": 27430 + }, + { + "epoch": 0.8469658620902525, + "grad_norm": 0.5408605337142944, + "learning_rate": 6.109802652506928e-05, + "loss": 0.0482, + "step": 27440 + }, + { + "epoch": 0.847274523118711, + "grad_norm": 0.32890963554382324, + "learning_rate": 6.107115442262291e-05, + "loss": 0.0487, + "step": 27450 + }, + { + "epoch": 0.8475831841471696, + "grad_norm": 0.3367066979408264, + "learning_rate": 6.104427895705214e-05, + "loss": 0.0486, + "step": 27460 + }, + { + "epoch": 0.8478918451756281, + "grad_norm": 0.3603740632534027, + "learning_rate": 6.101740013652103e-05, + "loss": 0.0423, + "step": 27470 + }, + { + "epoch": 0.8482005062040867, + "grad_norm": 0.31751707196235657, + "learning_rate": 6.099051796919465e-05, + "loss": 0.0437, + "step": 27480 + }, + { + "epoch": 0.8485091672325452, + "grad_norm": 0.4720967411994934, + "learning_rate": 6.096363246323911e-05, + "loss": 0.0455, + "step": 27490 + }, + { + "epoch": 0.8488178282610037, + "grad_norm": 0.381425142288208, + "learning_rate": 6.0936743626821504e-05, + "loss": 0.0399, + "step": 27500 + }, + { + "epoch": 0.8491264892894623, + "grad_norm": 0.3386598825454712, + "learning_rate": 6.090985146810996e-05, + "loss": 0.0455, + "step": 27510 + }, + { + "epoch": 0.8494351503179208, + "grad_norm": 0.2952543795108795, + "learning_rate": 6.088295599527357e-05, + "loss": 0.0504, + "step": 27520 + }, + { + "epoch": 0.8497438113463794, + "grad_norm": 0.389653742313385, + "learning_rate": 6.085605721648252e-05, + "loss": 0.042, + "step": 27530 + }, + { + "epoch": 0.8500524723748379, + "grad_norm": 0.3777545392513275, + "learning_rate": 6.082915513990792e-05, + "loss": 0.0453, + "step": 27540 + }, + { + "epoch": 0.8503611334032966, + "grad_norm": 1.9777567386627197, + "learning_rate": 6.080224977372192e-05, + "loss": 0.041, + "step": 27550 + }, + { + "epoch": 0.850669794431755, + "grad_norm": 0.41670092940330505, + "learning_rate": 6.0775341126097666e-05, + "loss": 0.0477, + "step": 27560 + }, + { + "epoch": 0.8509784554602136, + "grad_norm": 0.3955465853214264, + "learning_rate": 6.074842920520926e-05, + "loss": 0.0472, + "step": 27570 + }, + { + "epoch": 0.8512871164886722, + "grad_norm": 0.32566601037979126, + "learning_rate": 6.072151401923186e-05, + "loss": 0.0433, + "step": 27580 + }, + { + "epoch": 0.8515957775171307, + "grad_norm": 0.48497384786605835, + "learning_rate": 6.069459557634159e-05, + "loss": 0.0424, + "step": 27590 + }, + { + "epoch": 0.8519044385455893, + "grad_norm": 0.3660695254802704, + "learning_rate": 6.066767388471557e-05, + "loss": 0.048, + "step": 27600 + }, + { + "epoch": 0.8522130995740478, + "grad_norm": 0.41248244047164917, + "learning_rate": 6.064074895253188e-05, + "loss": 0.0438, + "step": 27610 + }, + { + "epoch": 0.8525217606025063, + "grad_norm": 0.47717130184173584, + "learning_rate": 6.061382078796961e-05, + "loss": 0.0437, + "step": 27620 + }, + { + "epoch": 0.8528304216309649, + "grad_norm": 0.3329247832298279, + "learning_rate": 6.0586889399208814e-05, + "loss": 0.0505, + "step": 27630 + }, + { + "epoch": 0.8531390826594234, + "grad_norm": 0.3460274934768677, + "learning_rate": 6.0559954794430565e-05, + "loss": 0.0464, + "step": 27640 + }, + { + "epoch": 0.853447743687882, + "grad_norm": 0.4000603258609772, + "learning_rate": 6.053301698181687e-05, + "loss": 0.0456, + "step": 27650 + }, + { + "epoch": 0.8537564047163405, + "grad_norm": 0.38701003789901733, + "learning_rate": 6.0506075969550725e-05, + "loss": 0.0461, + "step": 27660 + }, + { + "epoch": 0.854065065744799, + "grad_norm": 0.3171273171901703, + "learning_rate": 6.047913176581609e-05, + "loss": 0.0428, + "step": 27670 + }, + { + "epoch": 0.8543737267732576, + "grad_norm": 0.2988460958003998, + "learning_rate": 6.0452184378797904e-05, + "loss": 0.0456, + "step": 27680 + }, + { + "epoch": 0.8546823878017161, + "grad_norm": 0.3602847456932068, + "learning_rate": 6.042523381668209e-05, + "loss": 0.0474, + "step": 27690 + }, + { + "epoch": 0.8549910488301747, + "grad_norm": 0.44746914505958557, + "learning_rate": 6.03982800876555e-05, + "loss": 0.0437, + "step": 27700 + }, + { + "epoch": 0.8552997098586332, + "grad_norm": 0.6677365303039551, + "learning_rate": 6.0371323199905975e-05, + "loss": 0.0473, + "step": 27710 + }, + { + "epoch": 0.8556083708870917, + "grad_norm": 0.4019078314304352, + "learning_rate": 6.03443631616223e-05, + "loss": 0.0514, + "step": 27720 + }, + { + "epoch": 0.8559170319155504, + "grad_norm": 0.3837636113166809, + "learning_rate": 6.031739998099421e-05, + "loss": 0.0453, + "step": 27730 + }, + { + "epoch": 0.8562256929440089, + "grad_norm": 0.29995936155319214, + "learning_rate": 6.029043366621243e-05, + "loss": 0.0443, + "step": 27740 + }, + { + "epoch": 0.8565343539724675, + "grad_norm": 0.354483962059021, + "learning_rate": 6.0263464225468615e-05, + "loss": 0.0478, + "step": 27750 + }, + { + "epoch": 0.856843015000926, + "grad_norm": 0.2770715355873108, + "learning_rate": 6.023649166695534e-05, + "loss": 0.0462, + "step": 27760 + }, + { + "epoch": 0.8571516760293846, + "grad_norm": 0.27624279260635376, + "learning_rate": 6.0209515998866186e-05, + "loss": 0.0419, + "step": 27770 + }, + { + "epoch": 0.8574603370578431, + "grad_norm": 0.2849076986312866, + "learning_rate": 6.018253722939563e-05, + "loss": 0.0477, + "step": 27780 + }, + { + "epoch": 0.8577689980863016, + "grad_norm": 0.24588270485401154, + "learning_rate": 6.015555536673914e-05, + "loss": 0.0454, + "step": 27790 + }, + { + "epoch": 0.8580776591147602, + "grad_norm": 0.4099836051464081, + "learning_rate": 6.0128570419093054e-05, + "loss": 0.0497, + "step": 27800 + }, + { + "epoch": 0.8583863201432187, + "grad_norm": 0.37376052141189575, + "learning_rate": 6.010158239465471e-05, + "loss": 0.0449, + "step": 27810 + }, + { + "epoch": 0.8586949811716773, + "grad_norm": 0.3068268895149231, + "learning_rate": 6.007459130162235e-05, + "loss": 0.0475, + "step": 27820 + }, + { + "epoch": 0.8590036422001358, + "grad_norm": 0.35060399770736694, + "learning_rate": 6.004759714819516e-05, + "loss": 0.0477, + "step": 27830 + }, + { + "epoch": 0.8593123032285943, + "grad_norm": 0.30587896704673767, + "learning_rate": 6.002059994257323e-05, + "loss": 0.0392, + "step": 27840 + }, + { + "epoch": 0.8596209642570529, + "grad_norm": 0.3663739562034607, + "learning_rate": 5.999359969295764e-05, + "loss": 0.0445, + "step": 27850 + }, + { + "epoch": 0.8599296252855114, + "grad_norm": 0.3028263449668884, + "learning_rate": 5.9966596407550314e-05, + "loss": 0.0433, + "step": 27860 + }, + { + "epoch": 0.86023828631397, + "grad_norm": 0.538439154624939, + "learning_rate": 5.993959009455416e-05, + "loss": 0.0411, + "step": 27870 + }, + { + "epoch": 0.8605469473424285, + "grad_norm": 0.3309113085269928, + "learning_rate": 5.991258076217298e-05, + "loss": 0.0458, + "step": 27880 + }, + { + "epoch": 0.860855608370887, + "grad_norm": 0.3896249532699585, + "learning_rate": 5.988556841861147e-05, + "loss": 0.0478, + "step": 27890 + }, + { + "epoch": 0.8611642693993456, + "grad_norm": 0.325772762298584, + "learning_rate": 5.985855307207531e-05, + "loss": 0.0477, + "step": 27900 + }, + { + "epoch": 0.8614729304278042, + "grad_norm": 0.35123977065086365, + "learning_rate": 5.9831534730771e-05, + "loss": 0.0528, + "step": 27910 + }, + { + "epoch": 0.8617815914562628, + "grad_norm": 0.34089043736457825, + "learning_rate": 5.980451340290605e-05, + "loss": 0.0502, + "step": 27920 + }, + { + "epoch": 0.8620902524847213, + "grad_norm": 0.4435783922672272, + "learning_rate": 5.97774890966888e-05, + "loss": 0.0427, + "step": 27930 + }, + { + "epoch": 0.8623989135131799, + "grad_norm": 0.29753145575523376, + "learning_rate": 5.975046182032851e-05, + "loss": 0.0424, + "step": 27940 + }, + { + "epoch": 0.8627075745416384, + "grad_norm": 0.5111737251281738, + "learning_rate": 5.972343158203537e-05, + "loss": 0.0454, + "step": 27950 + }, + { + "epoch": 0.8630162355700969, + "grad_norm": 0.4020523726940155, + "learning_rate": 5.969639839002045e-05, + "loss": 0.0477, + "step": 27960 + }, + { + "epoch": 0.8633248965985555, + "grad_norm": 0.32688355445861816, + "learning_rate": 5.966936225249572e-05, + "loss": 0.0419, + "step": 27970 + }, + { + "epoch": 0.863633557627014, + "grad_norm": 0.3200508654117584, + "learning_rate": 5.9642323177674044e-05, + "loss": 0.0448, + "step": 27980 + }, + { + "epoch": 0.8639422186554726, + "grad_norm": 0.3833971917629242, + "learning_rate": 5.9615281173769154e-05, + "loss": 0.0457, + "step": 27990 + }, + { + "epoch": 0.8642508796839311, + "grad_norm": 0.2777441442012787, + "learning_rate": 5.958823624899574e-05, + "loss": 0.0481, + "step": 28000 + }, + { + "epoch": 0.8645595407123896, + "grad_norm": 0.3044670522212982, + "learning_rate": 5.956118841156933e-05, + "loss": 0.0484, + "step": 28010 + }, + { + "epoch": 0.8648682017408482, + "grad_norm": 0.37073594331741333, + "learning_rate": 5.953413766970631e-05, + "loss": 0.0468, + "step": 28020 + }, + { + "epoch": 0.8651768627693067, + "grad_norm": 0.38915300369262695, + "learning_rate": 5.9507084031624e-05, + "loss": 0.041, + "step": 28030 + }, + { + "epoch": 0.8654855237977653, + "grad_norm": 0.303122878074646, + "learning_rate": 5.948002750554058e-05, + "loss": 0.0487, + "step": 28040 + }, + { + "epoch": 0.8657941848262238, + "grad_norm": 0.37218499183654785, + "learning_rate": 5.9452968099675124e-05, + "loss": 0.0451, + "step": 28050 + }, + { + "epoch": 0.8661028458546823, + "grad_norm": 0.3105138838291168, + "learning_rate": 5.9425905822247527e-05, + "loss": 0.0479, + "step": 28060 + }, + { + "epoch": 0.866411506883141, + "grad_norm": 0.3393961787223816, + "learning_rate": 5.939884068147864e-05, + "loss": 0.0418, + "step": 28070 + }, + { + "epoch": 0.8667201679115994, + "grad_norm": 0.26048195362091064, + "learning_rate": 5.937177268559011e-05, + "loss": 0.0459, + "step": 28080 + }, + { + "epoch": 0.8670288289400581, + "grad_norm": 0.4134518504142761, + "learning_rate": 5.934470184280448e-05, + "loss": 0.0485, + "step": 28090 + }, + { + "epoch": 0.8673374899685166, + "grad_norm": 1.1515687704086304, + "learning_rate": 5.931762816134516e-05, + "loss": 0.0432, + "step": 28100 + }, + { + "epoch": 0.8676461509969752, + "grad_norm": 0.2679668068885803, + "learning_rate": 5.9290551649436434e-05, + "loss": 0.0463, + "step": 28110 + }, + { + "epoch": 0.8679548120254337, + "grad_norm": 0.3328106999397278, + "learning_rate": 5.9263472315303416e-05, + "loss": 0.0462, + "step": 28120 + }, + { + "epoch": 0.8682634730538922, + "grad_norm": 0.4287413954734802, + "learning_rate": 5.9236390167172096e-05, + "loss": 0.0478, + "step": 28130 + }, + { + "epoch": 0.8685721340823508, + "grad_norm": 0.35013505816459656, + "learning_rate": 5.920930521326932e-05, + "loss": 0.047, + "step": 28140 + }, + { + "epoch": 0.8688807951108093, + "grad_norm": 0.32043495774269104, + "learning_rate": 5.918221746182276e-05, + "loss": 0.0439, + "step": 28150 + }, + { + "epoch": 0.8691894561392679, + "grad_norm": 0.3425731360912323, + "learning_rate": 5.9155126921061e-05, + "loss": 0.0461, + "step": 28160 + }, + { + "epoch": 0.8694981171677264, + "grad_norm": 0.30409500002861023, + "learning_rate": 5.91280335992134e-05, + "loss": 0.0425, + "step": 28170 + }, + { + "epoch": 0.8698067781961849, + "grad_norm": 0.4040853679180145, + "learning_rate": 5.91009375045102e-05, + "loss": 0.0442, + "step": 28180 + }, + { + "epoch": 0.8701154392246435, + "grad_norm": 0.27512648701667786, + "learning_rate": 5.9073838645182476e-05, + "loss": 0.0449, + "step": 28190 + }, + { + "epoch": 0.870424100253102, + "grad_norm": 0.3514127731323242, + "learning_rate": 5.904673702946217e-05, + "loss": 0.0378, + "step": 28200 + }, + { + "epoch": 0.8707327612815606, + "grad_norm": 0.3057686984539032, + "learning_rate": 5.9019632665582004e-05, + "loss": 0.0413, + "step": 28210 + }, + { + "epoch": 0.8710414223100191, + "grad_norm": 0.33160412311553955, + "learning_rate": 5.899252556177559e-05, + "loss": 0.0429, + "step": 28220 + }, + { + "epoch": 0.8713500833384776, + "grad_norm": 0.4247969686985016, + "learning_rate": 5.896541572627735e-05, + "loss": 0.0432, + "step": 28230 + }, + { + "epoch": 0.8716587443669362, + "grad_norm": 0.4555995762348175, + "learning_rate": 5.893830316732253e-05, + "loss": 0.0436, + "step": 28240 + }, + { + "epoch": 0.8719674053953947, + "grad_norm": 0.27719488739967346, + "learning_rate": 5.8911187893147214e-05, + "loss": 0.0468, + "step": 28250 + }, + { + "epoch": 0.8722760664238534, + "grad_norm": 0.3634284436702728, + "learning_rate": 5.888406991198828e-05, + "loss": 0.0508, + "step": 28260 + }, + { + "epoch": 0.8725847274523119, + "grad_norm": 0.2403188943862915, + "learning_rate": 5.885694923208349e-05, + "loss": 0.0452, + "step": 28270 + }, + { + "epoch": 0.8728933884807705, + "grad_norm": 0.46208104491233826, + "learning_rate": 5.882982586167138e-05, + "loss": 0.0423, + "step": 28280 + }, + { + "epoch": 0.873202049509229, + "grad_norm": 0.33302903175354004, + "learning_rate": 5.880269980899131e-05, + "loss": 0.0442, + "step": 28290 + }, + { + "epoch": 0.8735107105376875, + "grad_norm": 0.31516948342323303, + "learning_rate": 5.8775571082283465e-05, + "loss": 0.0421, + "step": 28300 + }, + { + "epoch": 0.8738193715661461, + "grad_norm": 0.7202548384666443, + "learning_rate": 5.8748439689788824e-05, + "loss": 0.0504, + "step": 28310 + }, + { + "epoch": 0.8741280325946046, + "grad_norm": 0.3444448709487915, + "learning_rate": 5.87213056397492e-05, + "loss": 0.0424, + "step": 28320 + }, + { + "epoch": 0.8744366936230632, + "grad_norm": 0.2542075514793396, + "learning_rate": 5.869416894040719e-05, + "loss": 0.0418, + "step": 28330 + }, + { + "epoch": 0.8747453546515217, + "grad_norm": 0.2676968574523926, + "learning_rate": 5.866702960000621e-05, + "loss": 0.0467, + "step": 28340 + }, + { + "epoch": 0.8750540156799802, + "grad_norm": 0.33082646131515503, + "learning_rate": 5.863988762679048e-05, + "loss": 0.04, + "step": 28350 + }, + { + "epoch": 0.8753626767084388, + "grad_norm": 0.38717615604400635, + "learning_rate": 5.8612743029005e-05, + "loss": 0.048, + "step": 28360 + }, + { + "epoch": 0.8756713377368973, + "grad_norm": 0.35116350650787354, + "learning_rate": 5.858559581489561e-05, + "loss": 0.0414, + "step": 28370 + }, + { + "epoch": 0.8759799987653559, + "grad_norm": 0.5934101939201355, + "learning_rate": 5.85584459927089e-05, + "loss": 0.0456, + "step": 28380 + }, + { + "epoch": 0.8762886597938144, + "grad_norm": 0.29388853907585144, + "learning_rate": 5.853129357069227e-05, + "loss": 0.0454, + "step": 28390 + }, + { + "epoch": 0.8765973208222729, + "grad_norm": 0.27834245562553406, + "learning_rate": 5.8504138557093913e-05, + "loss": 0.0414, + "step": 28400 + }, + { + "epoch": 0.8769059818507315, + "grad_norm": 0.30189916491508484, + "learning_rate": 5.8476980960162784e-05, + "loss": 0.0416, + "step": 28410 + }, + { + "epoch": 0.87721464287919, + "grad_norm": 0.4485051929950714, + "learning_rate": 5.844982078814868e-05, + "loss": 0.041, + "step": 28420 + }, + { + "epoch": 0.8775233039076487, + "grad_norm": 0.27591559290885925, + "learning_rate": 5.842265804930211e-05, + "loss": 0.0422, + "step": 28430 + }, + { + "epoch": 0.8778319649361072, + "grad_norm": 0.2851393520832062, + "learning_rate": 5.839549275187444e-05, + "loss": 0.0412, + "step": 28440 + }, + { + "epoch": 0.8781406259645658, + "grad_norm": 0.3446996510028839, + "learning_rate": 5.836832490411771e-05, + "loss": 0.0434, + "step": 28450 + }, + { + "epoch": 0.8784492869930243, + "grad_norm": 0.3279985785484314, + "learning_rate": 5.834115451428485e-05, + "loss": 0.0397, + "step": 28460 + }, + { + "epoch": 0.8787579480214828, + "grad_norm": 0.3392941951751709, + "learning_rate": 5.831398159062946e-05, + "loss": 0.0447, + "step": 28470 + }, + { + "epoch": 0.8790666090499414, + "grad_norm": 0.7571160197257996, + "learning_rate": 5.828680614140599e-05, + "loss": 0.0476, + "step": 28480 + }, + { + "epoch": 0.8793752700783999, + "grad_norm": 0.462905615568161, + "learning_rate": 5.825962817486962e-05, + "loss": 0.0447, + "step": 28490 + }, + { + "epoch": 0.8796839311068585, + "grad_norm": 0.3491002023220062, + "learning_rate": 5.823244769927629e-05, + "loss": 0.0494, + "step": 28500 + }, + { + "epoch": 0.879992592135317, + "grad_norm": 0.3079887628555298, + "learning_rate": 5.8205264722882716e-05, + "loss": 0.0449, + "step": 28510 + }, + { + "epoch": 0.8803012531637755, + "grad_norm": 0.41958868503570557, + "learning_rate": 5.817807925394636e-05, + "loss": 0.0431, + "step": 28520 + }, + { + "epoch": 0.8806099141922341, + "grad_norm": 0.3049299716949463, + "learning_rate": 5.815089130072546e-05, + "loss": 0.0408, + "step": 28530 + }, + { + "epoch": 0.8809185752206926, + "grad_norm": 0.2721162736415863, + "learning_rate": 5.8123700871479e-05, + "loss": 0.0457, + "step": 28540 + }, + { + "epoch": 0.8812272362491512, + "grad_norm": 0.4423496723175049, + "learning_rate": 5.809650797446671e-05, + "loss": 0.0428, + "step": 28550 + }, + { + "epoch": 0.8815358972776097, + "grad_norm": 0.3532276153564453, + "learning_rate": 5.806931261794907e-05, + "loss": 0.0435, + "step": 28560 + }, + { + "epoch": 0.8818445583060682, + "grad_norm": 0.44893762469291687, + "learning_rate": 5.804211481018731e-05, + "loss": 0.0472, + "step": 28570 + }, + { + "epoch": 0.8821532193345268, + "grad_norm": 0.3494435250759125, + "learning_rate": 5.801491455944341e-05, + "loss": 0.0493, + "step": 28580 + }, + { + "epoch": 0.8824618803629853, + "grad_norm": 0.3010192811489105, + "learning_rate": 5.79877118739801e-05, + "loss": 0.0428, + "step": 28590 + }, + { + "epoch": 0.882770541391444, + "grad_norm": 0.2629449963569641, + "learning_rate": 5.7960506762060816e-05, + "loss": 0.0438, + "step": 28600 + }, + { + "epoch": 0.8830792024199025, + "grad_norm": 0.4024936258792877, + "learning_rate": 5.793329923194977e-05, + "loss": 0.0405, + "step": 28610 + }, + { + "epoch": 0.883387863448361, + "grad_norm": 0.4332776963710785, + "learning_rate": 5.790608929191187e-05, + "loss": 0.0446, + "step": 28620 + }, + { + "epoch": 0.8836965244768196, + "grad_norm": 0.30803099274635315, + "learning_rate": 5.78788769502128e-05, + "loss": 0.0468, + "step": 28630 + }, + { + "epoch": 0.8840051855052781, + "grad_norm": 0.5094947814941406, + "learning_rate": 5.785166221511894e-05, + "loss": 0.0425, + "step": 28640 + }, + { + "epoch": 0.8843138465337367, + "grad_norm": 0.5594860315322876, + "learning_rate": 5.7824445094897415e-05, + "loss": 0.0448, + "step": 28650 + }, + { + "epoch": 0.8846225075621952, + "grad_norm": 0.33174532651901245, + "learning_rate": 5.7797225597816065e-05, + "loss": 0.0421, + "step": 28660 + }, + { + "epoch": 0.8849311685906538, + "grad_norm": 0.26281899213790894, + "learning_rate": 5.777000373214345e-05, + "loss": 0.0488, + "step": 28670 + }, + { + "epoch": 0.8852398296191123, + "grad_norm": 0.32283103466033936, + "learning_rate": 5.774277950614885e-05, + "loss": 0.0391, + "step": 28680 + }, + { + "epoch": 0.8855484906475708, + "grad_norm": 0.3589443564414978, + "learning_rate": 5.771555292810227e-05, + "loss": 0.0462, + "step": 28690 + }, + { + "epoch": 0.8858571516760294, + "grad_norm": 0.29692941904067993, + "learning_rate": 5.768832400627444e-05, + "loss": 0.0448, + "step": 28700 + }, + { + "epoch": 0.8861658127044879, + "grad_norm": 0.37829869985580444, + "learning_rate": 5.7661092748936775e-05, + "loss": 0.0413, + "step": 28710 + }, + { + "epoch": 0.8864744737329465, + "grad_norm": 0.29395943880081177, + "learning_rate": 5.76338591643614e-05, + "loss": 0.0445, + "step": 28720 + }, + { + "epoch": 0.886783134761405, + "grad_norm": 0.3886600732803345, + "learning_rate": 5.760662326082118e-05, + "loss": 0.0404, + "step": 28730 + }, + { + "epoch": 0.8870917957898635, + "grad_norm": 0.3479345738887787, + "learning_rate": 5.757938504658965e-05, + "loss": 0.0417, + "step": 28740 + }, + { + "epoch": 0.8874004568183221, + "grad_norm": 0.3476465940475464, + "learning_rate": 5.755214452994107e-05, + "loss": 0.0484, + "step": 28750 + }, + { + "epoch": 0.8877091178467806, + "grad_norm": 0.4738902449607849, + "learning_rate": 5.752490171915039e-05, + "loss": 0.0455, + "step": 28760 + }, + { + "epoch": 0.8880177788752392, + "grad_norm": 0.3371782600879669, + "learning_rate": 5.749765662249324e-05, + "loss": 0.0435, + "step": 28770 + }, + { + "epoch": 0.8883264399036978, + "grad_norm": 0.351030170917511, + "learning_rate": 5.747040924824596e-05, + "loss": 0.0471, + "step": 28780 + }, + { + "epoch": 0.8886351009321563, + "grad_norm": 0.32855936884880066, + "learning_rate": 5.7443159604685613e-05, + "loss": 0.0415, + "step": 28790 + }, + { + "epoch": 0.8889437619606149, + "grad_norm": 0.32915639877319336, + "learning_rate": 5.74159077000899e-05, + "loss": 0.044, + "step": 28800 + }, + { + "epoch": 0.8892524229890734, + "grad_norm": 0.6434048414230347, + "learning_rate": 5.7388653542737235e-05, + "loss": 0.0476, + "step": 28810 + }, + { + "epoch": 0.889561084017532, + "grad_norm": 0.37660548090934753, + "learning_rate": 5.736139714090672e-05, + "loss": 0.0462, + "step": 28820 + }, + { + "epoch": 0.8898697450459905, + "grad_norm": 0.3622588813304901, + "learning_rate": 5.73341385028781e-05, + "loss": 0.0439, + "step": 28830 + }, + { + "epoch": 0.8901784060744491, + "grad_norm": 0.3316029906272888, + "learning_rate": 5.7306877636931855e-05, + "loss": 0.0419, + "step": 28840 + }, + { + "epoch": 0.8904870671029076, + "grad_norm": 0.27798840403556824, + "learning_rate": 5.7279614551349125e-05, + "loss": 0.0459, + "step": 28850 + }, + { + "epoch": 0.8907957281313661, + "grad_norm": 0.2704166769981384, + "learning_rate": 5.725234925441169e-05, + "loss": 0.0437, + "step": 28860 + }, + { + "epoch": 0.8911043891598247, + "grad_norm": 0.3942636251449585, + "learning_rate": 5.7225081754402044e-05, + "loss": 0.0508, + "step": 28870 + }, + { + "epoch": 0.8914130501882832, + "grad_norm": 0.3557836711406708, + "learning_rate": 5.7197812059603326e-05, + "loss": 0.0483, + "step": 28880 + }, + { + "epoch": 0.8917217112167418, + "grad_norm": 0.3376883268356323, + "learning_rate": 5.717054017829934e-05, + "loss": 0.0411, + "step": 28890 + }, + { + "epoch": 0.8920303722452003, + "grad_norm": 0.4585326910018921, + "learning_rate": 5.7143266118774584e-05, + "loss": 0.0463, + "step": 28900 + }, + { + "epoch": 0.8923390332736588, + "grad_norm": 0.3840252757072449, + "learning_rate": 5.711598988931418e-05, + "loss": 0.0477, + "step": 28910 + }, + { + "epoch": 0.8926476943021174, + "grad_norm": 0.2994535267353058, + "learning_rate": 5.7088711498203954e-05, + "loss": 0.0412, + "step": 28920 + }, + { + "epoch": 0.8929563553305759, + "grad_norm": 0.364156037569046, + "learning_rate": 5.706143095373033e-05, + "loss": 0.0462, + "step": 28930 + }, + { + "epoch": 0.8932650163590345, + "grad_norm": 0.3752012550830841, + "learning_rate": 5.703414826418042e-05, + "loss": 0.0426, + "step": 28940 + }, + { + "epoch": 0.893573677387493, + "grad_norm": 0.3572331666946411, + "learning_rate": 5.7006863437842007e-05, + "loss": 0.0474, + "step": 28950 + }, + { + "epoch": 0.8938823384159515, + "grad_norm": 0.3254645764827728, + "learning_rate": 5.697957648300348e-05, + "loss": 0.0469, + "step": 28960 + }, + { + "epoch": 0.8941909994444102, + "grad_norm": 0.3383522927761078, + "learning_rate": 5.695228740795391e-05, + "loss": 0.0456, + "step": 28970 + }, + { + "epoch": 0.8944996604728687, + "grad_norm": 0.2838481366634369, + "learning_rate": 5.6924996220982985e-05, + "loss": 0.0451, + "step": 28980 + }, + { + "epoch": 0.8948083215013273, + "grad_norm": 0.49552464485168457, + "learning_rate": 5.6897702930381045e-05, + "loss": 0.042, + "step": 28990 + }, + { + "epoch": 0.8951169825297858, + "grad_norm": 0.3587478995323181, + "learning_rate": 5.687040754443908e-05, + "loss": 0.0433, + "step": 29000 + }, + { + "epoch": 0.8954256435582444, + "grad_norm": 0.4216959476470947, + "learning_rate": 5.6843110071448725e-05, + "loss": 0.041, + "step": 29010 + }, + { + "epoch": 0.8957343045867029, + "grad_norm": 0.28676509857177734, + "learning_rate": 5.6815810519702194e-05, + "loss": 0.0447, + "step": 29020 + }, + { + "epoch": 0.8960429656151614, + "grad_norm": 0.31057703495025635, + "learning_rate": 5.6788508897492396e-05, + "loss": 0.04, + "step": 29030 + }, + { + "epoch": 0.89635162664362, + "grad_norm": 0.3218173086643219, + "learning_rate": 5.676120521311282e-05, + "loss": 0.0441, + "step": 29040 + }, + { + "epoch": 0.8966602876720785, + "grad_norm": 0.3173694908618927, + "learning_rate": 5.6733899474857634e-05, + "loss": 0.0486, + "step": 29050 + }, + { + "epoch": 0.8969689487005371, + "grad_norm": 0.32381966710090637, + "learning_rate": 5.670659169102157e-05, + "loss": 0.0456, + "step": 29060 + }, + { + "epoch": 0.8972776097289956, + "grad_norm": 0.28303274512290955, + "learning_rate": 5.6679281869900044e-05, + "loss": 0.0451, + "step": 29070 + }, + { + "epoch": 0.8975862707574541, + "grad_norm": 0.31933003664016724, + "learning_rate": 5.6651970019789045e-05, + "loss": 0.0434, + "step": 29080 + }, + { + "epoch": 0.8978949317859127, + "grad_norm": 0.3199852406978607, + "learning_rate": 5.662465614898519e-05, + "loss": 0.0397, + "step": 29090 + }, + { + "epoch": 0.8982035928143712, + "grad_norm": 0.3370893597602844, + "learning_rate": 5.6597340265785695e-05, + "loss": 0.0452, + "step": 29100 + }, + { + "epoch": 0.8985122538428298, + "grad_norm": 0.4242180585861206, + "learning_rate": 5.657002237848843e-05, + "loss": 0.045, + "step": 29110 + }, + { + "epoch": 0.8988209148712883, + "grad_norm": 0.3300336003303528, + "learning_rate": 5.654270249539183e-05, + "loss": 0.0482, + "step": 29120 + }, + { + "epoch": 0.8991295758997468, + "grad_norm": 0.2726519703865051, + "learning_rate": 5.651538062479498e-05, + "loss": 0.0447, + "step": 29130 + }, + { + "epoch": 0.8994382369282055, + "grad_norm": 0.3765130639076233, + "learning_rate": 5.648805677499751e-05, + "loss": 0.0466, + "step": 29140 + }, + { + "epoch": 0.899746897956664, + "grad_norm": 0.39101675152778625, + "learning_rate": 5.646073095429969e-05, + "loss": 0.0467, + "step": 29150 + }, + { + "epoch": 0.9000555589851226, + "grad_norm": 0.4450471103191376, + "learning_rate": 5.643340317100241e-05, + "loss": 0.0483, + "step": 29160 + }, + { + "epoch": 0.9003642200135811, + "grad_norm": 0.28885599970817566, + "learning_rate": 5.64060734334071e-05, + "loss": 0.0458, + "step": 29170 + }, + { + "epoch": 0.9006728810420397, + "grad_norm": 0.29720649123191833, + "learning_rate": 5.637874174981583e-05, + "loss": 0.0487, + "step": 29180 + }, + { + "epoch": 0.9009815420704982, + "grad_norm": 0.4071086049079895, + "learning_rate": 5.635140812853124e-05, + "loss": 0.043, + "step": 29190 + }, + { + "epoch": 0.9012902030989567, + "grad_norm": 0.4015354812145233, + "learning_rate": 5.6324072577856544e-05, + "loss": 0.0388, + "step": 29200 + }, + { + "epoch": 0.9015988641274153, + "grad_norm": 0.3552432954311371, + "learning_rate": 5.629673510609559e-05, + "loss": 0.0439, + "step": 29210 + }, + { + "epoch": 0.9019075251558738, + "grad_norm": 0.4177054464817047, + "learning_rate": 5.626939572155276e-05, + "loss": 0.0462, + "step": 29220 + }, + { + "epoch": 0.9022161861843324, + "grad_norm": 0.35952267050743103, + "learning_rate": 5.6242054432533054e-05, + "loss": 0.0466, + "step": 29230 + }, + { + "epoch": 0.9025248472127909, + "grad_norm": 0.2670317590236664, + "learning_rate": 5.621471124734201e-05, + "loss": 0.0457, + "step": 29240 + }, + { + "epoch": 0.9028335082412494, + "grad_norm": 0.36145302653312683, + "learning_rate": 5.6187366174285794e-05, + "loss": 0.0399, + "step": 29250 + }, + { + "epoch": 0.903142169269708, + "grad_norm": 0.384796142578125, + "learning_rate": 5.616001922167109e-05, + "loss": 0.0475, + "step": 29260 + }, + { + "epoch": 0.9034508302981665, + "grad_norm": 0.5983076691627502, + "learning_rate": 5.61326703978052e-05, + "loss": 0.0454, + "step": 29270 + }, + { + "epoch": 0.9037594913266251, + "grad_norm": 0.4995768368244171, + "learning_rate": 5.6105319710995964e-05, + "loss": 0.0452, + "step": 29280 + }, + { + "epoch": 0.9040681523550836, + "grad_norm": 0.4303486943244934, + "learning_rate": 5.60779671695518e-05, + "loss": 0.042, + "step": 29290 + }, + { + "epoch": 0.9043768133835421, + "grad_norm": 0.33252978324890137, + "learning_rate": 5.6050612781781684e-05, + "loss": 0.0451, + "step": 29300 + }, + { + "epoch": 0.9046854744120008, + "grad_norm": 0.35531631112098694, + "learning_rate": 5.602325655599516e-05, + "loss": 0.0406, + "step": 29310 + }, + { + "epoch": 0.9049941354404593, + "grad_norm": 0.37512561678886414, + "learning_rate": 5.599589850050234e-05, + "loss": 0.0454, + "step": 29320 + }, + { + "epoch": 0.9053027964689179, + "grad_norm": 0.28941744565963745, + "learning_rate": 5.5968538623613874e-05, + "loss": 0.0419, + "step": 29330 + }, + { + "epoch": 0.9056114574973764, + "grad_norm": 0.31387263536453247, + "learning_rate": 5.594117693364095e-05, + "loss": 0.0444, + "step": 29340 + }, + { + "epoch": 0.905920118525835, + "grad_norm": 1.033687710762024, + "learning_rate": 5.591381343889535e-05, + "loss": 0.0462, + "step": 29350 + }, + { + "epoch": 0.9062287795542935, + "grad_norm": 0.4765003025531769, + "learning_rate": 5.5886448147689355e-05, + "loss": 0.0435, + "step": 29360 + }, + { + "epoch": 0.906537440582752, + "grad_norm": 0.3047138750553131, + "learning_rate": 5.585908106833585e-05, + "loss": 0.0427, + "step": 29370 + }, + { + "epoch": 0.9068461016112106, + "grad_norm": 0.3842374384403229, + "learning_rate": 5.5831712209148226e-05, + "loss": 0.0415, + "step": 29380 + }, + { + "epoch": 0.9071547626396691, + "grad_norm": 1.317581057548523, + "learning_rate": 5.58043415784404e-05, + "loss": 0.047, + "step": 29390 + }, + { + "epoch": 0.9074634236681277, + "grad_norm": 0.3460563123226166, + "learning_rate": 5.577696918452686e-05, + "loss": 0.0435, + "step": 29400 + }, + { + "epoch": 0.9077720846965862, + "grad_norm": 0.3899767994880676, + "learning_rate": 5.5749595035722604e-05, + "loss": 0.0443, + "step": 29410 + }, + { + "epoch": 0.9080807457250447, + "grad_norm": 0.34228697419166565, + "learning_rate": 5.5722219140343193e-05, + "loss": 0.0469, + "step": 29420 + }, + { + "epoch": 0.9083894067535033, + "grad_norm": 0.30370190739631653, + "learning_rate": 5.56948415067047e-05, + "loss": 0.0508, + "step": 29430 + }, + { + "epoch": 0.9086980677819618, + "grad_norm": 0.28930607438087463, + "learning_rate": 5.5667462143123704e-05, + "loss": 0.0386, + "step": 29440 + }, + { + "epoch": 0.9090067288104204, + "grad_norm": 0.3235648274421692, + "learning_rate": 5.564008105791737e-05, + "loss": 0.0418, + "step": 29450 + }, + { + "epoch": 0.9093153898388789, + "grad_norm": 0.5323837995529175, + "learning_rate": 5.5612698259403316e-05, + "loss": 0.0475, + "step": 29460 + }, + { + "epoch": 0.9096240508673374, + "grad_norm": 0.3762740194797516, + "learning_rate": 5.5585313755899724e-05, + "loss": 0.0413, + "step": 29470 + }, + { + "epoch": 0.909932711895796, + "grad_norm": 0.4478108882904053, + "learning_rate": 5.5557927555725285e-05, + "loss": 0.0429, + "step": 29480 + }, + { + "epoch": 0.9102413729242546, + "grad_norm": 0.3746799826622009, + "learning_rate": 5.55305396671992e-05, + "loss": 0.0465, + "step": 29490 + }, + { + "epoch": 0.9105500339527132, + "grad_norm": 0.2850117087364197, + "learning_rate": 5.55031500986412e-05, + "loss": 0.0435, + "step": 29500 + }, + { + "epoch": 0.9108586949811717, + "grad_norm": 0.299791544675827, + "learning_rate": 5.547575885837149e-05, + "loss": 0.046, + "step": 29510 + }, + { + "epoch": 0.9111673560096302, + "grad_norm": 0.41186121106147766, + "learning_rate": 5.5448365954710825e-05, + "loss": 0.0459, + "step": 29520 + }, + { + "epoch": 0.9114760170380888, + "grad_norm": 0.40892714262008667, + "learning_rate": 5.5420971395980446e-05, + "loss": 0.0437, + "step": 29530 + }, + { + "epoch": 0.9117846780665473, + "grad_norm": 0.4256307780742645, + "learning_rate": 5.539357519050209e-05, + "loss": 0.045, + "step": 29540 + }, + { + "epoch": 0.9120933390950059, + "grad_norm": 0.3576734960079193, + "learning_rate": 5.536617734659799e-05, + "loss": 0.0419, + "step": 29550 + }, + { + "epoch": 0.9124020001234644, + "grad_norm": 0.3212689161300659, + "learning_rate": 5.533877787259091e-05, + "loss": 0.0442, + "step": 29560 + }, + { + "epoch": 0.912710661151923, + "grad_norm": 0.38708004355430603, + "learning_rate": 5.5311376776804044e-05, + "loss": 0.0405, + "step": 29570 + }, + { + "epoch": 0.9130193221803815, + "grad_norm": 0.3191330134868622, + "learning_rate": 5.528397406756118e-05, + "loss": 0.0488, + "step": 29580 + }, + { + "epoch": 0.91332798320884, + "grad_norm": 0.3218461573123932, + "learning_rate": 5.525656975318652e-05, + "loss": 0.042, + "step": 29590 + }, + { + "epoch": 0.9136366442372986, + "grad_norm": 0.33620885014533997, + "learning_rate": 5.522916384200474e-05, + "loss": 0.0441, + "step": 29600 + }, + { + "epoch": 0.9139453052657571, + "grad_norm": 0.2918839156627655, + "learning_rate": 5.520175634234106e-05, + "loss": 0.0446, + "step": 29610 + }, + { + "epoch": 0.9142539662942157, + "grad_norm": 0.596799910068512, + "learning_rate": 5.517434726252113e-05, + "loss": 0.0442, + "step": 29620 + }, + { + "epoch": 0.9145626273226742, + "grad_norm": 0.40128886699676514, + "learning_rate": 5.514693661087113e-05, + "loss": 0.0465, + "step": 29630 + }, + { + "epoch": 0.9148712883511327, + "grad_norm": 0.3556526005268097, + "learning_rate": 5.511952439571769e-05, + "loss": 0.0448, + "step": 29640 + }, + { + "epoch": 0.9151799493795914, + "grad_norm": 0.3361532390117645, + "learning_rate": 5.509211062538791e-05, + "loss": 0.0426, + "step": 29650 + }, + { + "epoch": 0.9154886104080499, + "grad_norm": 0.29160815477371216, + "learning_rate": 5.506469530820939e-05, + "loss": 0.0438, + "step": 29660 + }, + { + "epoch": 0.9157972714365085, + "grad_norm": 0.3942655920982361, + "learning_rate": 5.503727845251014e-05, + "loss": 0.0427, + "step": 29670 + }, + { + "epoch": 0.916105932464967, + "grad_norm": 0.4700351059436798, + "learning_rate": 5.50098600666187e-05, + "loss": 0.0451, + "step": 29680 + }, + { + "epoch": 0.9164145934934255, + "grad_norm": 0.3080258071422577, + "learning_rate": 5.498244015886406e-05, + "loss": 0.0439, + "step": 29690 + }, + { + "epoch": 0.9167232545218841, + "grad_norm": 0.7523461580276489, + "learning_rate": 5.495501873757565e-05, + "loss": 0.0447, + "step": 29700 + }, + { + "epoch": 0.9170319155503426, + "grad_norm": 0.2986522614955902, + "learning_rate": 5.492759581108336e-05, + "loss": 0.0445, + "step": 29710 + }, + { + "epoch": 0.9173405765788012, + "grad_norm": 0.3339863717556, + "learning_rate": 5.490017138771759e-05, + "loss": 0.0441, + "step": 29720 + }, + { + "epoch": 0.9176492376072597, + "grad_norm": 0.32742512226104736, + "learning_rate": 5.487274547580912e-05, + "loss": 0.0426, + "step": 29730 + }, + { + "epoch": 0.9179578986357183, + "grad_norm": 0.3204191327095032, + "learning_rate": 5.484531808368923e-05, + "loss": 0.0446, + "step": 29740 + }, + { + "epoch": 0.9182665596641768, + "grad_norm": 0.25667211413383484, + "learning_rate": 5.4817889219689656e-05, + "loss": 0.0467, + "step": 29750 + }, + { + "epoch": 0.9185752206926353, + "grad_norm": 0.32321155071258545, + "learning_rate": 5.4790458892142536e-05, + "loss": 0.0452, + "step": 29760 + }, + { + "epoch": 0.9188838817210939, + "grad_norm": 0.4357423782348633, + "learning_rate": 5.476302710938048e-05, + "loss": 0.0446, + "step": 29770 + }, + { + "epoch": 0.9191925427495524, + "grad_norm": 0.3892451524734497, + "learning_rate": 5.473559387973657e-05, + "loss": 0.0459, + "step": 29780 + }, + { + "epoch": 0.919501203778011, + "grad_norm": 0.30684584379196167, + "learning_rate": 5.470815921154425e-05, + "loss": 0.0426, + "step": 29790 + }, + { + "epoch": 0.9198098648064695, + "grad_norm": 0.7138813138008118, + "learning_rate": 5.468072311313749e-05, + "loss": 0.0472, + "step": 29800 + }, + { + "epoch": 0.920118525834928, + "grad_norm": 0.23287168145179749, + "learning_rate": 5.465328559285063e-05, + "loss": 0.0417, + "step": 29810 + }, + { + "epoch": 0.9204271868633866, + "grad_norm": 0.31076306104660034, + "learning_rate": 5.462584665901849e-05, + "loss": 0.0401, + "step": 29820 + }, + { + "epoch": 0.9207358478918451, + "grad_norm": 0.5264256596565247, + "learning_rate": 5.4598406319976235e-05, + "loss": 0.0426, + "step": 29830 + }, + { + "epoch": 0.9210445089203038, + "grad_norm": 0.338517427444458, + "learning_rate": 5.457096458405958e-05, + "loss": 0.0427, + "step": 29840 + }, + { + "epoch": 0.9213531699487623, + "grad_norm": 0.3312474489212036, + "learning_rate": 5.454352145960457e-05, + "loss": 0.045, + "step": 29850 + }, + { + "epoch": 0.9216618309772208, + "grad_norm": 0.37226298451423645, + "learning_rate": 5.4516076954947715e-05, + "loss": 0.0461, + "step": 29860 + }, + { + "epoch": 0.9219704920056794, + "grad_norm": 0.298871248960495, + "learning_rate": 5.448863107842591e-05, + "loss": 0.0443, + "step": 29870 + }, + { + "epoch": 0.9222791530341379, + "grad_norm": 0.3559145927429199, + "learning_rate": 5.446118383837651e-05, + "loss": 0.0455, + "step": 29880 + }, + { + "epoch": 0.9225878140625965, + "grad_norm": 0.3434011936187744, + "learning_rate": 5.443373524313722e-05, + "loss": 0.0474, + "step": 29890 + }, + { + "epoch": 0.922896475091055, + "grad_norm": 0.3559925854206085, + "learning_rate": 5.440628530104626e-05, + "loss": 0.0427, + "step": 29900 + }, + { + "epoch": 0.9232051361195136, + "grad_norm": 0.3649997413158417, + "learning_rate": 5.4378834020442146e-05, + "loss": 0.0433, + "step": 29910 + }, + { + "epoch": 0.9235137971479721, + "grad_norm": 0.2782806158065796, + "learning_rate": 5.4351381409663884e-05, + "loss": 0.0453, + "step": 29920 + }, + { + "epoch": 0.9238224581764306, + "grad_norm": 0.2799638509750366, + "learning_rate": 5.432392747705084e-05, + "loss": 0.0446, + "step": 29930 + }, + { + "epoch": 0.9241311192048892, + "grad_norm": 0.3221040368080139, + "learning_rate": 5.429647223094278e-05, + "loss": 0.0436, + "step": 29940 + }, + { + "epoch": 0.9244397802333477, + "grad_norm": 0.3695553243160248, + "learning_rate": 5.4269015679679924e-05, + "loss": 0.0442, + "step": 29950 + }, + { + "epoch": 0.9247484412618063, + "grad_norm": 0.3753278851509094, + "learning_rate": 5.424155783160281e-05, + "loss": 0.0417, + "step": 29960 + }, + { + "epoch": 0.9250571022902648, + "grad_norm": 0.3284580409526825, + "learning_rate": 5.4214098695052415e-05, + "loss": 0.0472, + "step": 29970 + }, + { + "epoch": 0.9253657633187233, + "grad_norm": 0.3393526077270508, + "learning_rate": 5.418663827837012e-05, + "loss": 0.0409, + "step": 29980 + }, + { + "epoch": 0.925674424347182, + "grad_norm": 0.4859170913696289, + "learning_rate": 5.415917658989763e-05, + "loss": 0.0413, + "step": 29990 + }, + { + "epoch": 0.9259830853756404, + "grad_norm": 0.5045261979103088, + "learning_rate": 5.413171363797713e-05, + "loss": 0.0428, + "step": 30000 + }, + { + "epoch": 0.9262917464040991, + "grad_norm": 0.29800671339035034, + "learning_rate": 5.4104249430951116e-05, + "loss": 0.0471, + "step": 30010 + }, + { + "epoch": 0.9266004074325576, + "grad_norm": 0.26931288838386536, + "learning_rate": 5.4076783977162494e-05, + "loss": 0.0491, + "step": 30020 + }, + { + "epoch": 0.9269090684610161, + "grad_norm": 0.3267909586429596, + "learning_rate": 5.4049317284954525e-05, + "loss": 0.0413, + "step": 30030 + }, + { + "epoch": 0.9272177294894747, + "grad_norm": 0.38585755228996277, + "learning_rate": 5.4021849362670884e-05, + "loss": 0.0449, + "step": 30040 + }, + { + "epoch": 0.9275263905179332, + "grad_norm": 0.2696080803871155, + "learning_rate": 5.3994380218655604e-05, + "loss": 0.0435, + "step": 30050 + }, + { + "epoch": 0.9278350515463918, + "grad_norm": 0.2935534715652466, + "learning_rate": 5.396690986125309e-05, + "loss": 0.0414, + "step": 30060 + }, + { + "epoch": 0.9281437125748503, + "grad_norm": 0.3677491545677185, + "learning_rate": 5.3939438298808075e-05, + "loss": 0.047, + "step": 30070 + }, + { + "epoch": 0.9284523736033089, + "grad_norm": 0.2937989830970764, + "learning_rate": 5.3911965539665744e-05, + "loss": 0.0383, + "step": 30080 + }, + { + "epoch": 0.9287610346317674, + "grad_norm": 0.2711375057697296, + "learning_rate": 5.388449159217156e-05, + "loss": 0.0467, + "step": 30090 + }, + { + "epoch": 0.9290696956602259, + "grad_norm": 0.37061285972595215, + "learning_rate": 5.3857016464671385e-05, + "loss": 0.0476, + "step": 30100 + }, + { + "epoch": 0.9293783566886845, + "grad_norm": 0.4195404350757599, + "learning_rate": 5.382954016551146e-05, + "loss": 0.0447, + "step": 30110 + }, + { + "epoch": 0.929687017717143, + "grad_norm": 0.2630762457847595, + "learning_rate": 5.380206270303835e-05, + "loss": 0.0404, + "step": 30120 + }, + { + "epoch": 0.9299956787456016, + "grad_norm": 0.3694535791873932, + "learning_rate": 5.377458408559897e-05, + "loss": 0.0502, + "step": 30130 + }, + { + "epoch": 0.9303043397740601, + "grad_norm": 0.42445942759513855, + "learning_rate": 5.374710432154061e-05, + "loss": 0.0462, + "step": 30140 + }, + { + "epoch": 0.9306130008025186, + "grad_norm": 0.9224503636360168, + "learning_rate": 5.3719623419210886e-05, + "loss": 0.0449, + "step": 30150 + }, + { + "epoch": 0.9309216618309772, + "grad_norm": 0.339272141456604, + "learning_rate": 5.3692141386957786e-05, + "loss": 0.0416, + "step": 30160 + }, + { + "epoch": 0.9312303228594357, + "grad_norm": 0.5020196437835693, + "learning_rate": 5.3664658233129616e-05, + "loss": 0.0435, + "step": 30170 + }, + { + "epoch": 0.9315389838878944, + "grad_norm": 0.5623682141304016, + "learning_rate": 5.363717396607504e-05, + "loss": 0.0462, + "step": 30180 + }, + { + "epoch": 0.9318476449163529, + "grad_norm": 0.5770459771156311, + "learning_rate": 5.360968859414305e-05, + "loss": 0.036, + "step": 30190 + }, + { + "epoch": 0.9321563059448114, + "grad_norm": 0.320471853017807, + "learning_rate": 5.358220212568295e-05, + "loss": 0.0435, + "step": 30200 + }, + { + "epoch": 0.93246496697327, + "grad_norm": 0.31702175736427307, + "learning_rate": 5.355471456904444e-05, + "loss": 0.0425, + "step": 30210 + }, + { + "epoch": 0.9327736280017285, + "grad_norm": 0.3980289697647095, + "learning_rate": 5.3527225932577495e-05, + "loss": 0.0394, + "step": 30220 + }, + { + "epoch": 0.9330822890301871, + "grad_norm": 0.37308621406555176, + "learning_rate": 5.349973622463246e-05, + "loss": 0.0461, + "step": 30230 + }, + { + "epoch": 0.9333909500586456, + "grad_norm": 0.3477040231227875, + "learning_rate": 5.3472245453559956e-05, + "loss": 0.0472, + "step": 30240 + }, + { + "epoch": 0.9336996110871042, + "grad_norm": 0.36341622471809387, + "learning_rate": 5.3444753627710955e-05, + "loss": 0.0443, + "step": 30250 + }, + { + "epoch": 0.9340082721155627, + "grad_norm": 0.3276067078113556, + "learning_rate": 5.341726075543676e-05, + "loss": 0.0404, + "step": 30260 + }, + { + "epoch": 0.9343169331440212, + "grad_norm": 0.2983282804489136, + "learning_rate": 5.338976684508898e-05, + "loss": 0.0441, + "step": 30270 + }, + { + "epoch": 0.9346255941724798, + "grad_norm": 0.311509370803833, + "learning_rate": 5.336227190501953e-05, + "loss": 0.0397, + "step": 30280 + }, + { + "epoch": 0.9349342552009383, + "grad_norm": 0.37553390860557556, + "learning_rate": 5.3334775943580664e-05, + "loss": 0.0414, + "step": 30290 + }, + { + "epoch": 0.9352429162293969, + "grad_norm": 0.3319146931171417, + "learning_rate": 5.330727896912491e-05, + "loss": 0.0466, + "step": 30300 + }, + { + "epoch": 0.9355515772578554, + "grad_norm": 12.256916999816895, + "learning_rate": 5.327978099000511e-05, + "loss": 0.0428, + "step": 30310 + }, + { + "epoch": 0.9358602382863139, + "grad_norm": 0.30648043751716614, + "learning_rate": 5.3252282014574465e-05, + "loss": 0.0411, + "step": 30320 + }, + { + "epoch": 0.9361688993147725, + "grad_norm": 0.5138270854949951, + "learning_rate": 5.322478205118641e-05, + "loss": 0.0426, + "step": 30330 + }, + { + "epoch": 0.936477560343231, + "grad_norm": 0.36600998044013977, + "learning_rate": 5.3197281108194704e-05, + "loss": 0.0421, + "step": 30340 + }, + { + "epoch": 0.9367862213716897, + "grad_norm": 0.48941928148269653, + "learning_rate": 5.316977919395342e-05, + "loss": 0.0449, + "step": 30350 + }, + { + "epoch": 0.9370948824001482, + "grad_norm": 0.32405611872673035, + "learning_rate": 5.314227631681691e-05, + "loss": 0.0434, + "step": 30360 + }, + { + "epoch": 0.9374035434286067, + "grad_norm": 0.29291409254074097, + "learning_rate": 5.311477248513982e-05, + "loss": 0.0434, + "step": 30370 + }, + { + "epoch": 0.9377122044570653, + "grad_norm": 0.36942940950393677, + "learning_rate": 5.30872677072771e-05, + "loss": 0.04, + "step": 30380 + }, + { + "epoch": 0.9380208654855238, + "grad_norm": 0.3262792229652405, + "learning_rate": 5.3059761991583954e-05, + "loss": 0.0419, + "step": 30390 + }, + { + "epoch": 0.9383295265139824, + "grad_norm": 0.550148606300354, + "learning_rate": 5.303225534641592e-05, + "loss": 0.0479, + "step": 30400 + }, + { + "epoch": 0.9386381875424409, + "grad_norm": 0.3144020736217499, + "learning_rate": 5.300474778012875e-05, + "loss": 0.043, + "step": 30410 + }, + { + "epoch": 0.9389468485708994, + "grad_norm": 0.4018910229206085, + "learning_rate": 5.297723930107855e-05, + "loss": 0.0445, + "step": 30420 + }, + { + "epoch": 0.939255509599358, + "grad_norm": 0.29896849393844604, + "learning_rate": 5.294972991762167e-05, + "loss": 0.0391, + "step": 30430 + }, + { + "epoch": 0.9395641706278165, + "grad_norm": 0.33691734075546265, + "learning_rate": 5.292221963811472e-05, + "loss": 0.0398, + "step": 30440 + }, + { + "epoch": 0.9398728316562751, + "grad_norm": 0.33612340688705444, + "learning_rate": 5.28947084709146e-05, + "loss": 0.0411, + "step": 30450 + }, + { + "epoch": 0.9401814926847336, + "grad_norm": 0.3610624670982361, + "learning_rate": 5.2867196424378465e-05, + "loss": 0.0442, + "step": 30460 + }, + { + "epoch": 0.9404901537131922, + "grad_norm": 0.31839442253112793, + "learning_rate": 5.2839683506863765e-05, + "loss": 0.0401, + "step": 30470 + }, + { + "epoch": 0.9407988147416507, + "grad_norm": 0.2568505108356476, + "learning_rate": 5.281216972672821e-05, + "loss": 0.041, + "step": 30480 + }, + { + "epoch": 0.9411074757701092, + "grad_norm": 0.9879403710365295, + "learning_rate": 5.278465509232973e-05, + "loss": 0.042, + "step": 30490 + }, + { + "epoch": 0.9414161367985678, + "grad_norm": 0.3259875476360321, + "learning_rate": 5.275713961202655e-05, + "loss": 0.0381, + "step": 30500 + }, + { + "epoch": 0.9417247978270263, + "grad_norm": 0.4151727557182312, + "learning_rate": 5.2729623294177165e-05, + "loss": 0.0417, + "step": 30510 + }, + { + "epoch": 0.942033458855485, + "grad_norm": 0.30514398217201233, + "learning_rate": 5.270210614714028e-05, + "loss": 0.0449, + "step": 30520 + }, + { + "epoch": 0.9423421198839435, + "grad_norm": 0.43450790643692017, + "learning_rate": 5.267458817927491e-05, + "loss": 0.0415, + "step": 30530 + }, + { + "epoch": 0.942650780912402, + "grad_norm": 0.35171449184417725, + "learning_rate": 5.264706939894026e-05, + "loss": 0.0466, + "step": 30540 + }, + { + "epoch": 0.9429594419408606, + "grad_norm": 0.30041563510894775, + "learning_rate": 5.261954981449584e-05, + "loss": 0.0413, + "step": 30550 + }, + { + "epoch": 0.9432681029693191, + "grad_norm": 0.5396047234535217, + "learning_rate": 5.2592029434301324e-05, + "loss": 0.0416, + "step": 30560 + }, + { + "epoch": 0.9435767639977777, + "grad_norm": 0.37945592403411865, + "learning_rate": 5.256450826671672e-05, + "loss": 0.0441, + "step": 30570 + }, + { + "epoch": 0.9438854250262362, + "grad_norm": 0.4643097519874573, + "learning_rate": 5.253698632010221e-05, + "loss": 0.0388, + "step": 30580 + }, + { + "epoch": 0.9441940860546947, + "grad_norm": 0.4011044204235077, + "learning_rate": 5.2509463602818246e-05, + "loss": 0.0426, + "step": 30590 + }, + { + "epoch": 0.9445027470831533, + "grad_norm": 0.4263392686843872, + "learning_rate": 5.248194012322549e-05, + "loss": 0.0447, + "step": 30600 + }, + { + "epoch": 0.9448114081116118, + "grad_norm": 0.42862850427627563, + "learning_rate": 5.245441588968486e-05, + "loss": 0.0464, + "step": 30610 + }, + { + "epoch": 0.9451200691400704, + "grad_norm": 0.3574641942977905, + "learning_rate": 5.242689091055748e-05, + "loss": 0.041, + "step": 30620 + }, + { + "epoch": 0.9454287301685289, + "grad_norm": 0.34994450211524963, + "learning_rate": 5.239936519420473e-05, + "loss": 0.0427, + "step": 30630 + }, + { + "epoch": 0.9457373911969875, + "grad_norm": 0.3785160183906555, + "learning_rate": 5.2371838748988175e-05, + "loss": 0.0402, + "step": 30640 + }, + { + "epoch": 0.946046052225446, + "grad_norm": 0.36894845962524414, + "learning_rate": 5.234431158326965e-05, + "loss": 0.0371, + "step": 30650 + }, + { + "epoch": 0.9463547132539045, + "grad_norm": 0.4303695857524872, + "learning_rate": 5.231678370541115e-05, + "loss": 0.041, + "step": 30660 + }, + { + "epoch": 0.9466633742823631, + "grad_norm": 0.41653311252593994, + "learning_rate": 5.228925512377495e-05, + "loss": 0.0428, + "step": 30670 + }, + { + "epoch": 0.9469720353108216, + "grad_norm": 0.35278603434562683, + "learning_rate": 5.2261725846723465e-05, + "loss": 0.0456, + "step": 30680 + }, + { + "epoch": 0.9472806963392802, + "grad_norm": 0.32176172733306885, + "learning_rate": 5.22341958826194e-05, + "loss": 0.0402, + "step": 30690 + }, + { + "epoch": 0.9475893573677387, + "grad_norm": 0.3121320307254791, + "learning_rate": 5.22066652398256e-05, + "loss": 0.0433, + "step": 30700 + }, + { + "epoch": 0.9478980183961973, + "grad_norm": 0.31201496720314026, + "learning_rate": 5.2179133926705185e-05, + "loss": 0.0472, + "step": 30710 + }, + { + "epoch": 0.9482066794246559, + "grad_norm": 0.3261583745479584, + "learning_rate": 5.215160195162141e-05, + "loss": 0.0449, + "step": 30720 + }, + { + "epoch": 0.9485153404531144, + "grad_norm": 0.45973503589630127, + "learning_rate": 5.212406932293776e-05, + "loss": 0.0465, + "step": 30730 + }, + { + "epoch": 0.948824001481573, + "grad_norm": 0.6261411905288696, + "learning_rate": 5.209653604901795e-05, + "loss": 0.0423, + "step": 30740 + }, + { + "epoch": 0.9491326625100315, + "grad_norm": 0.40169382095336914, + "learning_rate": 5.206900213822584e-05, + "loss": 0.0461, + "step": 30750 + }, + { + "epoch": 0.94944132353849, + "grad_norm": 0.3277452290058136, + "learning_rate": 5.204146759892551e-05, + "loss": 0.0429, + "step": 30760 + }, + { + "epoch": 0.9497499845669486, + "grad_norm": 0.2835303544998169, + "learning_rate": 5.2013932439481216e-05, + "loss": 0.0414, + "step": 30770 + }, + { + "epoch": 0.9500586455954071, + "grad_norm": 0.5361688733100891, + "learning_rate": 5.198639666825743e-05, + "loss": 0.0431, + "step": 30780 + }, + { + "epoch": 0.9503673066238657, + "grad_norm": 0.25042954087257385, + "learning_rate": 5.195886029361877e-05, + "loss": 0.0417, + "step": 30790 + }, + { + "epoch": 0.9506759676523242, + "grad_norm": 0.37503674626350403, + "learning_rate": 5.193132332393009e-05, + "loss": 0.0434, + "step": 30800 + }, + { + "epoch": 0.9509846286807828, + "grad_norm": 0.39268723130226135, + "learning_rate": 5.1903785767556376e-05, + "loss": 0.0414, + "step": 30810 + }, + { + "epoch": 0.9512932897092413, + "grad_norm": 0.31936994194984436, + "learning_rate": 5.187624763286282e-05, + "loss": 0.0441, + "step": 30820 + }, + { + "epoch": 0.9516019507376998, + "grad_norm": 0.4150725305080414, + "learning_rate": 5.184870892821475e-05, + "loss": 0.0446, + "step": 30830 + }, + { + "epoch": 0.9519106117661584, + "grad_norm": 0.5603005886077881, + "learning_rate": 5.182116966197773e-05, + "loss": 0.0403, + "step": 30840 + }, + { + "epoch": 0.9522192727946169, + "grad_norm": 0.339088499546051, + "learning_rate": 5.1793629842517466e-05, + "loss": 0.0415, + "step": 30850 + }, + { + "epoch": 0.9525279338230755, + "grad_norm": 0.29580897092819214, + "learning_rate": 5.17660894781998e-05, + "loss": 0.0429, + "step": 30860 + }, + { + "epoch": 0.952836594851534, + "grad_norm": 0.35037460923194885, + "learning_rate": 5.173854857739079e-05, + "loss": 0.0396, + "step": 30870 + }, + { + "epoch": 0.9531452558799925, + "grad_norm": 0.32339048385620117, + "learning_rate": 5.171100714845661e-05, + "loss": 0.0429, + "step": 30880 + }, + { + "epoch": 0.9534539169084512, + "grad_norm": 0.2808716893196106, + "learning_rate": 5.1683465199763646e-05, + "loss": 0.0436, + "step": 30890 + }, + { + "epoch": 0.9537625779369097, + "grad_norm": 0.39069536328315735, + "learning_rate": 5.16559227396784e-05, + "loss": 0.0498, + "step": 30900 + }, + { + "epoch": 0.9540712389653683, + "grad_norm": 0.37926360964775085, + "learning_rate": 5.1628379776567556e-05, + "loss": 0.0407, + "step": 30910 + }, + { + "epoch": 0.9543798999938268, + "grad_norm": 0.40377822518348694, + "learning_rate": 5.160083631879792e-05, + "loss": 0.0379, + "step": 30920 + }, + { + "epoch": 0.9546885610222853, + "grad_norm": 0.37369129061698914, + "learning_rate": 5.1573292374736484e-05, + "loss": 0.0423, + "step": 30930 + }, + { + "epoch": 0.9549972220507439, + "grad_norm": 0.27170923352241516, + "learning_rate": 5.1545747952750356e-05, + "loss": 0.041, + "step": 30940 + }, + { + "epoch": 0.9553058830792024, + "grad_norm": 0.34872838854789734, + "learning_rate": 5.151820306120682e-05, + "loss": 0.047, + "step": 30950 + }, + { + "epoch": 0.955614544107661, + "grad_norm": 0.4153873324394226, + "learning_rate": 5.149065770847328e-05, + "loss": 0.0487, + "step": 30960 + }, + { + "epoch": 0.9559232051361195, + "grad_norm": 0.3298870325088501, + "learning_rate": 5.1463111902917297e-05, + "loss": 0.0391, + "step": 30970 + }, + { + "epoch": 0.9562318661645781, + "grad_norm": 0.3310655355453491, + "learning_rate": 5.143556565290654e-05, + "loss": 0.0421, + "step": 30980 + }, + { + "epoch": 0.9565405271930366, + "grad_norm": 0.3250986933708191, + "learning_rate": 5.140801896680882e-05, + "loss": 0.0461, + "step": 30990 + }, + { + "epoch": 0.9568491882214951, + "grad_norm": 0.45335742831230164, + "learning_rate": 5.1380471852992144e-05, + "loss": 0.0411, + "step": 31000 + }, + { + "epoch": 0.9571578492499537, + "grad_norm": 0.5249902009963989, + "learning_rate": 5.135292431982457e-05, + "loss": 0.0455, + "step": 31010 + }, + { + "epoch": 0.9574665102784122, + "grad_norm": 0.5014258027076721, + "learning_rate": 5.1325376375674294e-05, + "loss": 0.0428, + "step": 31020 + }, + { + "epoch": 0.9577751713068708, + "grad_norm": 0.3220387399196625, + "learning_rate": 5.129782802890968e-05, + "loss": 0.045, + "step": 31030 + }, + { + "epoch": 0.9580838323353293, + "grad_norm": 0.30069637298583984, + "learning_rate": 5.127027928789916e-05, + "loss": 0.0419, + "step": 31040 + }, + { + "epoch": 0.9583924933637878, + "grad_norm": 0.33702248334884644, + "learning_rate": 5.124273016101135e-05, + "loss": 0.041, + "step": 31050 + }, + { + "epoch": 0.9587011543922465, + "grad_norm": 0.28744807839393616, + "learning_rate": 5.121518065661492e-05, + "loss": 0.045, + "step": 31060 + }, + { + "epoch": 0.959009815420705, + "grad_norm": 0.3281744718551636, + "learning_rate": 5.11876307830787e-05, + "loss": 0.0389, + "step": 31070 + }, + { + "epoch": 0.9593184764491636, + "grad_norm": 0.4250691533088684, + "learning_rate": 5.1160080548771596e-05, + "loss": 0.0471, + "step": 31080 + }, + { + "epoch": 0.9596271374776221, + "grad_norm": 0.36627185344696045, + "learning_rate": 5.1132529962062656e-05, + "loss": 0.0412, + "step": 31090 + }, + { + "epoch": 0.9599357985060806, + "grad_norm": 0.28060972690582275, + "learning_rate": 5.110497903132101e-05, + "loss": 0.0397, + "step": 31100 + }, + { + "epoch": 0.9602444595345392, + "grad_norm": 0.31853342056274414, + "learning_rate": 5.107742776491592e-05, + "loss": 0.045, + "step": 31110 + }, + { + "epoch": 0.9605531205629977, + "grad_norm": 0.2996561527252197, + "learning_rate": 5.104987617121673e-05, + "loss": 0.0446, + "step": 31120 + }, + { + "epoch": 0.9608617815914563, + "grad_norm": 0.3248111307621002, + "learning_rate": 5.102232425859287e-05, + "loss": 0.0453, + "step": 31130 + }, + { + "epoch": 0.9611704426199148, + "grad_norm": 0.8564184308052063, + "learning_rate": 5.09947720354139e-05, + "loss": 0.0429, + "step": 31140 + }, + { + "epoch": 0.9614791036483733, + "grad_norm": 0.42689669132232666, + "learning_rate": 5.096721951004942e-05, + "loss": 0.0412, + "step": 31150 + }, + { + "epoch": 0.9617877646768319, + "grad_norm": 0.30197495222091675, + "learning_rate": 5.0939666690869227e-05, + "loss": 0.0392, + "step": 31160 + }, + { + "epoch": 0.9620964257052904, + "grad_norm": 0.32939594984054565, + "learning_rate": 5.0912113586243096e-05, + "loss": 0.044, + "step": 31170 + }, + { + "epoch": 0.962405086733749, + "grad_norm": 0.4311906695365906, + "learning_rate": 5.0884560204540935e-05, + "loss": 0.0441, + "step": 31180 + }, + { + "epoch": 0.9627137477622075, + "grad_norm": 0.38220417499542236, + "learning_rate": 5.0857006554132736e-05, + "loss": 0.0482, + "step": 31190 + }, + { + "epoch": 0.9630224087906661, + "grad_norm": 0.39090871810913086, + "learning_rate": 5.0829452643388575e-05, + "loss": 0.0434, + "step": 31200 + }, + { + "epoch": 0.9633310698191246, + "grad_norm": 0.40267348289489746, + "learning_rate": 5.08018984806786e-05, + "loss": 0.0407, + "step": 31210 + }, + { + "epoch": 0.9636397308475831, + "grad_norm": 0.4104400873184204, + "learning_rate": 5.0774344074373036e-05, + "loss": 0.0431, + "step": 31220 + }, + { + "epoch": 0.9639483918760418, + "grad_norm": 0.36083030700683594, + "learning_rate": 5.07467894328422e-05, + "loss": 0.0399, + "step": 31230 + }, + { + "epoch": 0.9642570529045003, + "grad_norm": 0.39024513959884644, + "learning_rate": 5.0719234564456454e-05, + "loss": 0.0403, + "step": 31240 + }, + { + "epoch": 0.9645657139329589, + "grad_norm": 0.3284365236759186, + "learning_rate": 5.0691679477586216e-05, + "loss": 0.0441, + "step": 31250 + }, + { + "epoch": 0.9648743749614174, + "grad_norm": 0.3168134391307831, + "learning_rate": 5.0664124180602035e-05, + "loss": 0.0415, + "step": 31260 + }, + { + "epoch": 0.9651830359898759, + "grad_norm": 0.32054710388183594, + "learning_rate": 5.063656868187447e-05, + "loss": 0.0447, + "step": 31270 + }, + { + "epoch": 0.9654916970183345, + "grad_norm": 0.24517224729061127, + "learning_rate": 5.060901298977413e-05, + "loss": 0.0395, + "step": 31280 + }, + { + "epoch": 0.965800358046793, + "grad_norm": 0.27733445167541504, + "learning_rate": 5.0581457112671725e-05, + "loss": 0.0455, + "step": 31290 + }, + { + "epoch": 0.9661090190752516, + "grad_norm": 0.46085143089294434, + "learning_rate": 5.0553901058938016e-05, + "loss": 0.0434, + "step": 31300 + }, + { + "epoch": 0.9664176801037101, + "grad_norm": 0.5854632258415222, + "learning_rate": 5.052634483694377e-05, + "loss": 0.0445, + "step": 31310 + }, + { + "epoch": 0.9667263411321686, + "grad_norm": 0.31725332140922546, + "learning_rate": 5.049878845505988e-05, + "loss": 0.0427, + "step": 31320 + }, + { + "epoch": 0.9670350021606272, + "grad_norm": 0.3605499863624573, + "learning_rate": 5.047123192165721e-05, + "loss": 0.046, + "step": 31330 + }, + { + "epoch": 0.9673436631890857, + "grad_norm": 0.6255589723587036, + "learning_rate": 5.0443675245106735e-05, + "loss": 0.0468, + "step": 31340 + }, + { + "epoch": 0.9676523242175443, + "grad_norm": 0.3971309959888458, + "learning_rate": 5.0416118433779426e-05, + "loss": 0.0411, + "step": 31350 + }, + { + "epoch": 0.9679609852460028, + "grad_norm": 0.2711468040943146, + "learning_rate": 5.038856149604633e-05, + "loss": 0.0395, + "step": 31360 + }, + { + "epoch": 0.9682696462744614, + "grad_norm": 0.4449102282524109, + "learning_rate": 5.03610044402785e-05, + "loss": 0.0489, + "step": 31370 + }, + { + "epoch": 0.9685783073029199, + "grad_norm": 0.39955466985702515, + "learning_rate": 5.033344727484707e-05, + "loss": 0.0395, + "step": 31380 + }, + { + "epoch": 0.9688869683313784, + "grad_norm": 0.490445077419281, + "learning_rate": 5.030589000812315e-05, + "loss": 0.0457, + "step": 31390 + }, + { + "epoch": 0.969195629359837, + "grad_norm": 0.3542807400226593, + "learning_rate": 5.027833264847793e-05, + "loss": 0.0426, + "step": 31400 + }, + { + "epoch": 0.9695042903882956, + "grad_norm": 0.4045597016811371, + "learning_rate": 5.025077520428258e-05, + "loss": 0.0481, + "step": 31410 + }, + { + "epoch": 0.9698129514167542, + "grad_norm": 0.5839622020721436, + "learning_rate": 5.022321768390837e-05, + "loss": 0.043, + "step": 31420 + }, + { + "epoch": 0.9701216124452127, + "grad_norm": 0.33892765641212463, + "learning_rate": 5.0195660095726516e-05, + "loss": 0.0411, + "step": 31430 + }, + { + "epoch": 0.9704302734736712, + "grad_norm": 0.4135657250881195, + "learning_rate": 5.016810244810829e-05, + "loss": 0.0435, + "step": 31440 + }, + { + "epoch": 0.9707389345021298, + "grad_norm": 0.2439427673816681, + "learning_rate": 5.0140544749424976e-05, + "loss": 0.04, + "step": 31450 + }, + { + "epoch": 0.9710475955305883, + "grad_norm": 0.33692070841789246, + "learning_rate": 5.0112987008047874e-05, + "loss": 0.0379, + "step": 31460 + }, + { + "epoch": 0.9713562565590469, + "grad_norm": 0.38941100239753723, + "learning_rate": 5.008542923234831e-05, + "loss": 0.0446, + "step": 31470 + }, + { + "epoch": 0.9716649175875054, + "grad_norm": 0.43601319193840027, + "learning_rate": 5.00578714306976e-05, + "loss": 0.0432, + "step": 31480 + }, + { + "epoch": 0.9719735786159639, + "grad_norm": 0.3523854613304138, + "learning_rate": 5.0030313611467084e-05, + "loss": 0.0398, + "step": 31490 + }, + { + "epoch": 0.9722822396444225, + "grad_norm": 0.4517101049423218, + "learning_rate": 5.0002755783028074e-05, + "loss": 0.0383, + "step": 31500 + }, + { + "epoch": 0.972590900672881, + "grad_norm": 0.3478144109249115, + "learning_rate": 4.997519795375194e-05, + "loss": 0.0423, + "step": 31510 + }, + { + "epoch": 0.9728995617013396, + "grad_norm": 1.7986563444137573, + "learning_rate": 4.9947640132010016e-05, + "loss": 0.0432, + "step": 31520 + }, + { + "epoch": 0.9732082227297981, + "grad_norm": 0.3720586895942688, + "learning_rate": 4.9920082326173625e-05, + "loss": 0.0436, + "step": 31530 + }, + { + "epoch": 0.9735168837582567, + "grad_norm": 0.8379294872283936, + "learning_rate": 4.9892524544614114e-05, + "loss": 0.0402, + "step": 31540 + }, + { + "epoch": 0.9738255447867152, + "grad_norm": 0.31363779306411743, + "learning_rate": 4.986496679570283e-05, + "loss": 0.0451, + "step": 31550 + }, + { + "epoch": 0.9741342058151737, + "grad_norm": 0.4090811014175415, + "learning_rate": 4.983740908781105e-05, + "loss": 0.0412, + "step": 31560 + }, + { + "epoch": 0.9744428668436323, + "grad_norm": 0.49521535634994507, + "learning_rate": 4.9809851429310116e-05, + "loss": 0.0426, + "step": 31570 + }, + { + "epoch": 0.9747515278720909, + "grad_norm": 0.49352410435676575, + "learning_rate": 4.9782293828571275e-05, + "loss": 0.0396, + "step": 31580 + }, + { + "epoch": 0.9750601889005495, + "grad_norm": 0.32964658737182617, + "learning_rate": 4.9754736293965846e-05, + "loss": 0.0396, + "step": 31590 + }, + { + "epoch": 0.975368849929008, + "grad_norm": 0.32488712668418884, + "learning_rate": 4.972717883386502e-05, + "loss": 0.0416, + "step": 31600 + }, + { + "epoch": 0.9756775109574665, + "grad_norm": 0.3441922962665558, + "learning_rate": 4.9699621456640075e-05, + "loss": 0.0411, + "step": 31610 + }, + { + "epoch": 0.9759861719859251, + "grad_norm": 0.3138032555580139, + "learning_rate": 4.9672064170662214e-05, + "loss": 0.0454, + "step": 31620 + }, + { + "epoch": 0.9762948330143836, + "grad_norm": 0.32948577404022217, + "learning_rate": 4.9644506984302583e-05, + "loss": 0.0456, + "step": 31630 + }, + { + "epoch": 0.9766034940428422, + "grad_norm": 0.42727765440940857, + "learning_rate": 4.9616949905932356e-05, + "loss": 0.0413, + "step": 31640 + }, + { + "epoch": 0.9769121550713007, + "grad_norm": 0.4042387902736664, + "learning_rate": 4.9589392943922615e-05, + "loss": 0.0442, + "step": 31650 + }, + { + "epoch": 0.9772208160997592, + "grad_norm": 0.376977801322937, + "learning_rate": 4.956183610664447e-05, + "loss": 0.0396, + "step": 31660 + }, + { + "epoch": 0.9775294771282178, + "grad_norm": 0.3246733546257019, + "learning_rate": 4.9534279402468945e-05, + "loss": 0.0422, + "step": 31670 + }, + { + "epoch": 0.9778381381566763, + "grad_norm": 0.32511255145072937, + "learning_rate": 4.9506722839767036e-05, + "loss": 0.043, + "step": 31680 + }, + { + "epoch": 0.9781467991851349, + "grad_norm": 0.2912295460700989, + "learning_rate": 4.947916642690972e-05, + "loss": 0.0395, + "step": 31690 + }, + { + "epoch": 0.9784554602135934, + "grad_norm": 0.3353446424007416, + "learning_rate": 4.9451610172267874e-05, + "loss": 0.0437, + "step": 31700 + }, + { + "epoch": 0.978764121242052, + "grad_norm": 0.35015052556991577, + "learning_rate": 4.9424054084212376e-05, + "loss": 0.0434, + "step": 31710 + }, + { + "epoch": 0.9790727822705105, + "grad_norm": 0.36940932273864746, + "learning_rate": 4.939649817111407e-05, + "loss": 0.0393, + "step": 31720 + }, + { + "epoch": 0.979381443298969, + "grad_norm": 0.3835791051387787, + "learning_rate": 4.936894244134365e-05, + "loss": 0.0412, + "step": 31730 + }, + { + "epoch": 0.9796901043274276, + "grad_norm": 0.49766838550567627, + "learning_rate": 4.9341386903271886e-05, + "loss": 0.0406, + "step": 31740 + }, + { + "epoch": 0.9799987653558861, + "grad_norm": 0.5290414690971375, + "learning_rate": 4.931383156526936e-05, + "loss": 0.0428, + "step": 31750 + }, + { + "epoch": 0.9803074263843448, + "grad_norm": 0.4486440122127533, + "learning_rate": 4.92862764357067e-05, + "loss": 0.0406, + "step": 31760 + }, + { + "epoch": 0.9806160874128033, + "grad_norm": 0.44061726331710815, + "learning_rate": 4.925872152295443e-05, + "loss": 0.0434, + "step": 31770 + }, + { + "epoch": 0.9809247484412618, + "grad_norm": 0.3227224349975586, + "learning_rate": 4.923116683538296e-05, + "loss": 0.0432, + "step": 31780 + }, + { + "epoch": 0.9812334094697204, + "grad_norm": 0.32517725229263306, + "learning_rate": 4.920361238136273e-05, + "loss": 0.0509, + "step": 31790 + }, + { + "epoch": 0.9815420704981789, + "grad_norm": 0.28828755021095276, + "learning_rate": 4.9176058169264014e-05, + "loss": 0.0408, + "step": 31800 + }, + { + "epoch": 0.9818507315266375, + "grad_norm": 0.32731935381889343, + "learning_rate": 4.9148504207457074e-05, + "loss": 0.0412, + "step": 31810 + }, + { + "epoch": 0.982159392555096, + "grad_norm": 0.333244651556015, + "learning_rate": 4.912095050431208e-05, + "loss": 0.0451, + "step": 31820 + }, + { + "epoch": 0.9824680535835545, + "grad_norm": 0.3547690212726593, + "learning_rate": 4.909339706819911e-05, + "loss": 0.0423, + "step": 31830 + }, + { + "epoch": 0.9827767146120131, + "grad_norm": 0.47242745757102966, + "learning_rate": 4.906584390748819e-05, + "loss": 0.0433, + "step": 31840 + }, + { + "epoch": 0.9830853756404716, + "grad_norm": 0.3088567554950714, + "learning_rate": 4.9038291030549195e-05, + "loss": 0.0469, + "step": 31850 + }, + { + "epoch": 0.9833940366689302, + "grad_norm": 0.3300608992576599, + "learning_rate": 4.9010738445751995e-05, + "loss": 0.0444, + "step": 31860 + }, + { + "epoch": 0.9837026976973887, + "grad_norm": 0.2873293459415436, + "learning_rate": 4.8983186161466364e-05, + "loss": 0.0414, + "step": 31870 + }, + { + "epoch": 0.9840113587258473, + "grad_norm": 0.31612247228622437, + "learning_rate": 4.89556341860619e-05, + "loss": 0.0417, + "step": 31880 + }, + { + "epoch": 0.9843200197543058, + "grad_norm": 0.405504047870636, + "learning_rate": 4.892808252790822e-05, + "loss": 0.0442, + "step": 31890 + }, + { + "epoch": 0.9846286807827643, + "grad_norm": 0.3190388083457947, + "learning_rate": 4.890053119537475e-05, + "loss": 0.0443, + "step": 31900 + }, + { + "epoch": 0.9849373418112229, + "grad_norm": 0.4336208403110504, + "learning_rate": 4.887298019683087e-05, + "loss": 0.0412, + "step": 31910 + }, + { + "epoch": 0.9852460028396814, + "grad_norm": 0.5149561762809753, + "learning_rate": 4.884542954064587e-05, + "loss": 0.0382, + "step": 31920 + }, + { + "epoch": 0.9855546638681401, + "grad_norm": 0.33692389726638794, + "learning_rate": 4.881787923518887e-05, + "loss": 0.0416, + "step": 31930 + }, + { + "epoch": 0.9858633248965986, + "grad_norm": 0.3327774405479431, + "learning_rate": 4.879032928882896e-05, + "loss": 0.043, + "step": 31940 + }, + { + "epoch": 0.9861719859250571, + "grad_norm": 0.4774811863899231, + "learning_rate": 4.876277970993505e-05, + "loss": 0.0494, + "step": 31950 + }, + { + "epoch": 0.9864806469535157, + "grad_norm": 0.3063490092754364, + "learning_rate": 4.873523050687602e-05, + "loss": 0.0385, + "step": 31960 + }, + { + "epoch": 0.9867893079819742, + "grad_norm": 0.3397417962551117, + "learning_rate": 4.870768168802056e-05, + "loss": 0.0477, + "step": 31970 + }, + { + "epoch": 0.9870979690104328, + "grad_norm": 0.3093611001968384, + "learning_rate": 4.868013326173728e-05, + "loss": 0.044, + "step": 31980 + }, + { + "epoch": 0.9874066300388913, + "grad_norm": 0.3731456696987152, + "learning_rate": 4.865258523639468e-05, + "loss": 0.0382, + "step": 31990 + }, + { + "epoch": 0.9877152910673498, + "grad_norm": 0.27688726782798767, + "learning_rate": 4.862503762036109e-05, + "loss": 0.0403, + "step": 32000 + }, + { + "epoch": 0.9880239520958084, + "grad_norm": 0.24425843358039856, + "learning_rate": 4.859749042200478e-05, + "loss": 0.0434, + "step": 32010 + }, + { + "epoch": 0.9883326131242669, + "grad_norm": 0.36838987469673157, + "learning_rate": 4.856994364969384e-05, + "loss": 0.0427, + "step": 32020 + }, + { + "epoch": 0.9886412741527255, + "grad_norm": 0.3704480528831482, + "learning_rate": 4.854239731179625e-05, + "loss": 0.048, + "step": 32030 + }, + { + "epoch": 0.988949935181184, + "grad_norm": 0.45937737822532654, + "learning_rate": 4.85148514166799e-05, + "loss": 0.0437, + "step": 32040 + }, + { + "epoch": 0.9892585962096425, + "grad_norm": 0.408692330121994, + "learning_rate": 4.8487305972712456e-05, + "loss": 0.0453, + "step": 32050 + }, + { + "epoch": 0.9895672572381011, + "grad_norm": 0.31302985548973083, + "learning_rate": 4.8459760988261526e-05, + "loss": 0.0396, + "step": 32060 + }, + { + "epoch": 0.9898759182665596, + "grad_norm": 0.42580437660217285, + "learning_rate": 4.843221647169453e-05, + "loss": 0.0401, + "step": 32070 + }, + { + "epoch": 0.9901845792950182, + "grad_norm": 0.32734179496765137, + "learning_rate": 4.840467243137878e-05, + "loss": 0.0433, + "step": 32080 + }, + { + "epoch": 0.9904932403234767, + "grad_norm": 0.3346024453639984, + "learning_rate": 4.837712887568143e-05, + "loss": 0.0415, + "step": 32090 + }, + { + "epoch": 0.9908019013519354, + "grad_norm": 0.3437190353870392, + "learning_rate": 4.8349585812969464e-05, + "loss": 0.0446, + "step": 32100 + }, + { + "epoch": 0.9911105623803939, + "grad_norm": 0.35777029395103455, + "learning_rate": 4.8322043251609775e-05, + "loss": 0.0417, + "step": 32110 + }, + { + "epoch": 0.9914192234088524, + "grad_norm": 0.41786324977874756, + "learning_rate": 4.8294501199969015e-05, + "loss": 0.0428, + "step": 32120 + }, + { + "epoch": 0.991727884437311, + "grad_norm": 0.32540392875671387, + "learning_rate": 4.826695966641376e-05, + "loss": 0.0444, + "step": 32130 + }, + { + "epoch": 0.9920365454657695, + "grad_norm": 0.3696131110191345, + "learning_rate": 4.823941865931043e-05, + "loss": 0.0432, + "step": 32140 + }, + { + "epoch": 0.9923452064942281, + "grad_norm": 0.280173122882843, + "learning_rate": 4.82118781870252e-05, + "loss": 0.0402, + "step": 32150 + }, + { + "epoch": 0.9926538675226866, + "grad_norm": 0.43352189660072327, + "learning_rate": 4.8184338257924185e-05, + "loss": 0.0435, + "step": 32160 + }, + { + "epoch": 0.9929625285511451, + "grad_norm": 0.3405858278274536, + "learning_rate": 4.815679888037324e-05, + "loss": 0.0411, + "step": 32170 + }, + { + "epoch": 0.9932711895796037, + "grad_norm": 0.31317439675331116, + "learning_rate": 4.8129260062738135e-05, + "loss": 0.0391, + "step": 32180 + }, + { + "epoch": 0.9935798506080622, + "grad_norm": 0.30403369665145874, + "learning_rate": 4.810172181338445e-05, + "loss": 0.0437, + "step": 32190 + }, + { + "epoch": 0.9938885116365208, + "grad_norm": 0.5545105338096619, + "learning_rate": 4.807418414067753e-05, + "loss": 0.0424, + "step": 32200 + }, + { + "epoch": 0.9941971726649793, + "grad_norm": 0.39930039644241333, + "learning_rate": 4.804664705298264e-05, + "loss": 0.041, + "step": 32210 + }, + { + "epoch": 0.9945058336934378, + "grad_norm": 0.42405033111572266, + "learning_rate": 4.80191105586648e-05, + "loss": 0.044, + "step": 32220 + }, + { + "epoch": 0.9948144947218964, + "grad_norm": 0.509781002998352, + "learning_rate": 4.799157466608886e-05, + "loss": 0.0432, + "step": 32230 + }, + { + "epoch": 0.9951231557503549, + "grad_norm": 0.36470940709114075, + "learning_rate": 4.796403938361951e-05, + "loss": 0.0439, + "step": 32240 + }, + { + "epoch": 0.9954318167788135, + "grad_norm": 0.3517559766769409, + "learning_rate": 4.793650471962123e-05, + "loss": 0.0407, + "step": 32250 + }, + { + "epoch": 0.995740477807272, + "grad_norm": 0.3204779624938965, + "learning_rate": 4.790897068245835e-05, + "loss": 0.0414, + "step": 32260 + }, + { + "epoch": 0.9960491388357307, + "grad_norm": 0.27765437960624695, + "learning_rate": 4.7881437280494954e-05, + "loss": 0.0433, + "step": 32270 + }, + { + "epoch": 0.9963577998641892, + "grad_norm": 0.42432737350463867, + "learning_rate": 4.7853904522094965e-05, + "loss": 0.0407, + "step": 32280 + }, + { + "epoch": 0.9966664608926477, + "grad_norm": 0.2788452208042145, + "learning_rate": 4.782637241562215e-05, + "loss": 0.0427, + "step": 32290 + }, + { + "epoch": 0.9969751219211063, + "grad_norm": 0.32891902327537537, + "learning_rate": 4.779884096943997e-05, + "loss": 0.0451, + "step": 32300 + }, + { + "epoch": 0.9972837829495648, + "grad_norm": 0.3891068696975708, + "learning_rate": 4.777131019191182e-05, + "loss": 0.0419, + "step": 32310 + }, + { + "epoch": 0.9975924439780234, + "grad_norm": 0.25987687706947327, + "learning_rate": 4.774378009140076e-05, + "loss": 0.046, + "step": 32320 + }, + { + "epoch": 0.9979011050064819, + "grad_norm": 0.27243444323539734, + "learning_rate": 4.7716250676269735e-05, + "loss": 0.04, + "step": 32330 + }, + { + "epoch": 0.9982097660349404, + "grad_norm": 0.3720300495624542, + "learning_rate": 4.7688721954881485e-05, + "loss": 0.0453, + "step": 32340 + }, + { + "epoch": 0.998518427063399, + "grad_norm": 0.2571966350078583, + "learning_rate": 4.7661193935598446e-05, + "loss": 0.0422, + "step": 32350 + }, + { + "epoch": 0.9988270880918575, + "grad_norm": 0.354481041431427, + "learning_rate": 4.763366662678296e-05, + "loss": 0.0396, + "step": 32360 + }, + { + "epoch": 0.9991357491203161, + "grad_norm": 0.2661072015762329, + "learning_rate": 4.7606140036797064e-05, + "loss": 0.0397, + "step": 32370 + }, + { + "epoch": 0.9994444101487746, + "grad_norm": 0.2724267542362213, + "learning_rate": 4.7578614174002614e-05, + "loss": 0.0468, + "step": 32380 + }, + { + "epoch": 0.9997530711772331, + "grad_norm": 0.4484831392765045, + "learning_rate": 4.755108904676125e-05, + "loss": 0.0411, + "step": 32390 + }, + { + "epoch": 1.0000617322056917, + "grad_norm": 0.2899266183376312, + "learning_rate": 4.752356466343436e-05, + "loss": 0.0438, + "step": 32400 + }, + { + "epoch": 1.0003703932341503, + "grad_norm": 0.24274657666683197, + "learning_rate": 4.7496041032383174e-05, + "loss": 0.0413, + "step": 32410 + }, + { + "epoch": 1.0006790542626087, + "grad_norm": 0.339788019657135, + "learning_rate": 4.746851816196858e-05, + "loss": 0.0469, + "step": 32420 + }, + { + "epoch": 1.0009877152910673, + "grad_norm": 0.32147181034088135, + "learning_rate": 4.744099606055135e-05, + "loss": 0.0444, + "step": 32430 + }, + { + "epoch": 1.001296376319526, + "grad_norm": 0.3945755660533905, + "learning_rate": 4.741347473649193e-05, + "loss": 0.0428, + "step": 32440 + }, + { + "epoch": 1.0016050373479843, + "grad_norm": 0.3550388216972351, + "learning_rate": 4.738595419815058e-05, + "loss": 0.0425, + "step": 32450 + }, + { + "epoch": 1.001913698376443, + "grad_norm": 0.3291845917701721, + "learning_rate": 4.7358434453887365e-05, + "loss": 0.0384, + "step": 32460 + }, + { + "epoch": 1.0022223594049016, + "grad_norm": 0.24626219272613525, + "learning_rate": 4.7330915512061976e-05, + "loss": 0.0446, + "step": 32470 + }, + { + "epoch": 1.0025310204333602, + "grad_norm": 0.350561261177063, + "learning_rate": 4.730339738103402e-05, + "loss": 0.041, + "step": 32480 + }, + { + "epoch": 1.0028396814618186, + "grad_norm": 0.34045663475990295, + "learning_rate": 4.727588006916271e-05, + "loss": 0.0437, + "step": 32490 + }, + { + "epoch": 1.0031483424902772, + "grad_norm": 0.37994375824928284, + "learning_rate": 4.724836358480711e-05, + "loss": 0.0465, + "step": 32500 + }, + { + "epoch": 1.0034570035187358, + "grad_norm": 0.47691577672958374, + "learning_rate": 4.722084793632601e-05, + "loss": 0.0434, + "step": 32510 + }, + { + "epoch": 1.0037656645471942, + "grad_norm": 0.35054221749305725, + "learning_rate": 4.719333313207792e-05, + "loss": 0.0416, + "step": 32520 + }, + { + "epoch": 1.0040743255756528, + "grad_norm": 0.3103416860103607, + "learning_rate": 4.716581918042114e-05, + "loss": 0.0447, + "step": 32530 + }, + { + "epoch": 1.0043829866041114, + "grad_norm": 0.3531530499458313, + "learning_rate": 4.7138306089713636e-05, + "loss": 0.0413, + "step": 32540 + }, + { + "epoch": 1.00469164763257, + "grad_norm": 0.295986533164978, + "learning_rate": 4.7110793868313183e-05, + "loss": 0.0446, + "step": 32550 + }, + { + "epoch": 1.0050003086610284, + "grad_norm": 0.31152287125587463, + "learning_rate": 4.708328252457729e-05, + "loss": 0.0429, + "step": 32560 + }, + { + "epoch": 1.005308969689487, + "grad_norm": 0.3659031093120575, + "learning_rate": 4.7055772066863135e-05, + "loss": 0.0439, + "step": 32570 + }, + { + "epoch": 1.0056176307179456, + "grad_norm": 0.2347102165222168, + "learning_rate": 4.702826250352771e-05, + "loss": 0.0403, + "step": 32580 + }, + { + "epoch": 1.005926291746404, + "grad_norm": 0.8204516172409058, + "learning_rate": 4.7000753842927653e-05, + "loss": 0.0398, + "step": 32590 + }, + { + "epoch": 1.0062349527748626, + "grad_norm": 0.3103034496307373, + "learning_rate": 4.6973246093419384e-05, + "loss": 0.041, + "step": 32600 + }, + { + "epoch": 1.0065436138033212, + "grad_norm": 0.40419432520866394, + "learning_rate": 4.694573926335906e-05, + "loss": 0.0464, + "step": 32610 + }, + { + "epoch": 1.0068522748317796, + "grad_norm": 0.30611392855644226, + "learning_rate": 4.6918233361102476e-05, + "loss": 0.041, + "step": 32620 + }, + { + "epoch": 1.0071609358602382, + "grad_norm": 0.344615638256073, + "learning_rate": 4.689072839500525e-05, + "loss": 0.0427, + "step": 32630 + }, + { + "epoch": 1.0074695968886969, + "grad_norm": 0.4059871733188629, + "learning_rate": 4.6863224373422635e-05, + "loss": 0.0406, + "step": 32640 + }, + { + "epoch": 1.0077782579171555, + "grad_norm": 0.3668518662452698, + "learning_rate": 4.683572130470962e-05, + "loss": 0.0449, + "step": 32650 + }, + { + "epoch": 1.0080869189456139, + "grad_norm": 0.46490126848220825, + "learning_rate": 4.680821919722094e-05, + "loss": 0.0382, + "step": 32660 + }, + { + "epoch": 1.0083955799740725, + "grad_norm": 0.3214133679866791, + "learning_rate": 4.6780718059310975e-05, + "loss": 0.0398, + "step": 32670 + }, + { + "epoch": 1.008704241002531, + "grad_norm": 0.25887295603752136, + "learning_rate": 4.675321789933389e-05, + "loss": 0.0391, + "step": 32680 + }, + { + "epoch": 1.0090129020309895, + "grad_norm": 0.3355702757835388, + "learning_rate": 4.6725718725643464e-05, + "loss": 0.0415, + "step": 32690 + }, + { + "epoch": 1.009321563059448, + "grad_norm": 0.30599382519721985, + "learning_rate": 4.669822054659323e-05, + "loss": 0.0471, + "step": 32700 + }, + { + "epoch": 1.0096302240879067, + "grad_norm": 0.3301965296268463, + "learning_rate": 4.667072337053644e-05, + "loss": 0.0419, + "step": 32710 + }, + { + "epoch": 1.0099388851163653, + "grad_norm": 0.3462929427623749, + "learning_rate": 4.6643227205825965e-05, + "loss": 0.0388, + "step": 32720 + }, + { + "epoch": 1.0102475461448237, + "grad_norm": 0.3201686441898346, + "learning_rate": 4.6615732060814454e-05, + "loss": 0.0437, + "step": 32730 + }, + { + "epoch": 1.0105562071732823, + "grad_norm": 0.37677690386772156, + "learning_rate": 4.658823794385417e-05, + "loss": 0.0382, + "step": 32740 + }, + { + "epoch": 1.010864868201741, + "grad_norm": 0.3279290497303009, + "learning_rate": 4.6560744863297115e-05, + "loss": 0.0419, + "step": 32750 + }, + { + "epoch": 1.0111735292301993, + "grad_norm": 0.255957692861557, + "learning_rate": 4.653325282749498e-05, + "loss": 0.0442, + "step": 32760 + }, + { + "epoch": 1.011482190258658, + "grad_norm": 0.3176335394382477, + "learning_rate": 4.6505761844799075e-05, + "loss": 0.0441, + "step": 32770 + }, + { + "epoch": 1.0117908512871165, + "grad_norm": 0.38015270233154297, + "learning_rate": 4.647827192356048e-05, + "loss": 0.0402, + "step": 32780 + }, + { + "epoch": 1.012099512315575, + "grad_norm": 0.30330371856689453, + "learning_rate": 4.645078307212989e-05, + "loss": 0.0455, + "step": 32790 + }, + { + "epoch": 1.0124081733440335, + "grad_norm": 0.32751989364624023, + "learning_rate": 4.642329529885768e-05, + "loss": 0.0373, + "step": 32800 + }, + { + "epoch": 1.0127168343724922, + "grad_norm": 0.364966481924057, + "learning_rate": 4.639580861209393e-05, + "loss": 0.0407, + "step": 32810 + }, + { + "epoch": 1.0130254954009508, + "grad_norm": 0.3200511932373047, + "learning_rate": 4.636832302018835e-05, + "loss": 0.0405, + "step": 32820 + }, + { + "epoch": 1.0133341564294092, + "grad_norm": 0.3933258652687073, + "learning_rate": 4.6340838531490365e-05, + "loss": 0.0371, + "step": 32830 + }, + { + "epoch": 1.0136428174578678, + "grad_norm": 0.30283793807029724, + "learning_rate": 4.6313355154349e-05, + "loss": 0.0426, + "step": 32840 + }, + { + "epoch": 1.0139514784863264, + "grad_norm": 0.384559690952301, + "learning_rate": 4.6285872897113025e-05, + "loss": 0.0434, + "step": 32850 + }, + { + "epoch": 1.0142601395147848, + "grad_norm": 0.34961435198783875, + "learning_rate": 4.625839176813077e-05, + "loss": 0.0385, + "step": 32860 + }, + { + "epoch": 1.0145688005432434, + "grad_norm": 0.2890254855155945, + "learning_rate": 4.623091177575031e-05, + "loss": 0.0435, + "step": 32870 + }, + { + "epoch": 1.014877461571702, + "grad_norm": 0.33843448758125305, + "learning_rate": 4.620343292831936e-05, + "loss": 0.041, + "step": 32880 + }, + { + "epoch": 1.0151861226001606, + "grad_norm": 0.3102676272392273, + "learning_rate": 4.6175955234185206e-05, + "loss": 0.0425, + "step": 32890 + }, + { + "epoch": 1.015494783628619, + "grad_norm": 0.30977851152420044, + "learning_rate": 4.614847870169492e-05, + "loss": 0.0417, + "step": 32900 + }, + { + "epoch": 1.0158034446570776, + "grad_norm": 0.36633041501045227, + "learning_rate": 4.612100333919509e-05, + "loss": 0.0405, + "step": 32910 + }, + { + "epoch": 1.0161121056855362, + "grad_norm": 0.334589421749115, + "learning_rate": 4.609352915503202e-05, + "loss": 0.0381, + "step": 32920 + }, + { + "epoch": 1.0164207667139946, + "grad_norm": 0.3323107063770294, + "learning_rate": 4.606605615755166e-05, + "loss": 0.0382, + "step": 32930 + }, + { + "epoch": 1.0167294277424532, + "grad_norm": 0.2842155396938324, + "learning_rate": 4.6038584355099576e-05, + "loss": 0.0393, + "step": 32940 + }, + { + "epoch": 1.0170380887709118, + "grad_norm": 0.4209918975830078, + "learning_rate": 4.6011113756020964e-05, + "loss": 0.0401, + "step": 32950 + }, + { + "epoch": 1.0173467497993702, + "grad_norm": 0.34483665227890015, + "learning_rate": 4.598364436866066e-05, + "loss": 0.0409, + "step": 32960 + }, + { + "epoch": 1.0176554108278288, + "grad_norm": 0.4414425492286682, + "learning_rate": 4.595617620136316e-05, + "loss": 0.0435, + "step": 32970 + }, + { + "epoch": 1.0179640718562875, + "grad_norm": 0.31847792863845825, + "learning_rate": 4.592870926247257e-05, + "loss": 0.0363, + "step": 32980 + }, + { + "epoch": 1.018272732884746, + "grad_norm": 0.34626638889312744, + "learning_rate": 4.5901243560332594e-05, + "loss": 0.0404, + "step": 32990 + }, + { + "epoch": 1.0185813939132045, + "grad_norm": 0.5467674136161804, + "learning_rate": 4.587377910328662e-05, + "loss": 0.0398, + "step": 33000 + }, + { + "epoch": 1.018890054941663, + "grad_norm": 0.37161919474601746, + "learning_rate": 4.5846315899677586e-05, + "loss": 0.0438, + "step": 33010 + }, + { + "epoch": 1.0191987159701217, + "grad_norm": 0.35011929273605347, + "learning_rate": 4.5818853957848114e-05, + "loss": 0.0392, + "step": 33020 + }, + { + "epoch": 1.01950737699858, + "grad_norm": 0.3572058081626892, + "learning_rate": 4.579139328614043e-05, + "loss": 0.0428, + "step": 33030 + }, + { + "epoch": 1.0198160380270387, + "grad_norm": 0.3872525095939636, + "learning_rate": 4.576393389289633e-05, + "loss": 0.0383, + "step": 33040 + }, + { + "epoch": 1.0201246990554973, + "grad_norm": 0.3400253653526306, + "learning_rate": 4.573647578645728e-05, + "loss": 0.0414, + "step": 33050 + }, + { + "epoch": 1.0204333600839557, + "grad_norm": 0.3836219608783722, + "learning_rate": 4.57090189751643e-05, + "loss": 0.0397, + "step": 33060 + }, + { + "epoch": 1.0207420211124143, + "grad_norm": 0.2316940724849701, + "learning_rate": 4.568156346735806e-05, + "loss": 0.0367, + "step": 33070 + }, + { + "epoch": 1.021050682140873, + "grad_norm": 0.3230094909667969, + "learning_rate": 4.565410927137882e-05, + "loss": 0.043, + "step": 33080 + }, + { + "epoch": 1.0213593431693315, + "grad_norm": 0.2893522083759308, + "learning_rate": 4.562665639556644e-05, + "loss": 0.0414, + "step": 33090 + }, + { + "epoch": 1.02166800419779, + "grad_norm": 0.444203644990921, + "learning_rate": 4.559920484826037e-05, + "loss": 0.0395, + "step": 33100 + }, + { + "epoch": 1.0219766652262485, + "grad_norm": 0.267144113779068, + "learning_rate": 4.5571754637799665e-05, + "loss": 0.0392, + "step": 33110 + }, + { + "epoch": 1.0222853262547071, + "grad_norm": 0.22843311727046967, + "learning_rate": 4.554430577252298e-05, + "loss": 0.0355, + "step": 33120 + }, + { + "epoch": 1.0225939872831655, + "grad_norm": 0.29683539271354675, + "learning_rate": 4.551685826076858e-05, + "loss": 0.0404, + "step": 33130 + }, + { + "epoch": 1.0229026483116241, + "grad_norm": 0.40672022104263306, + "learning_rate": 4.5489412110874246e-05, + "loss": 0.0377, + "step": 33140 + }, + { + "epoch": 1.0232113093400828, + "grad_norm": 0.421816349029541, + "learning_rate": 4.5461967331177444e-05, + "loss": 0.0444, + "step": 33150 + }, + { + "epoch": 1.0235199703685414, + "grad_norm": 0.289650022983551, + "learning_rate": 4.5434523930015115e-05, + "loss": 0.0383, + "step": 33160 + }, + { + "epoch": 1.0238286313969998, + "grad_norm": 0.45391520857810974, + "learning_rate": 4.540708191572388e-05, + "loss": 0.0399, + "step": 33170 + }, + { + "epoch": 1.0241372924254584, + "grad_norm": 0.32526493072509766, + "learning_rate": 4.537964129663991e-05, + "loss": 0.0375, + "step": 33180 + }, + { + "epoch": 1.024445953453917, + "grad_norm": 0.5574694871902466, + "learning_rate": 4.535220208109889e-05, + "loss": 0.0404, + "step": 33190 + }, + { + "epoch": 1.0247546144823754, + "grad_norm": 0.226614311337471, + "learning_rate": 4.5324764277436194e-05, + "loss": 0.0392, + "step": 33200 + }, + { + "epoch": 1.025063275510834, + "grad_norm": 0.4085023105144501, + "learning_rate": 4.529732789398664e-05, + "loss": 0.0462, + "step": 33210 + }, + { + "epoch": 1.0253719365392926, + "grad_norm": 0.3085777759552002, + "learning_rate": 4.526989293908472e-05, + "loss": 0.041, + "step": 33220 + }, + { + "epoch": 1.0256805975677512, + "grad_norm": 0.3613167107105255, + "learning_rate": 4.524245942106442e-05, + "loss": 0.0428, + "step": 33230 + }, + { + "epoch": 1.0259892585962096, + "grad_norm": 0.2884882092475891, + "learning_rate": 4.5215027348259345e-05, + "loss": 0.0434, + "step": 33240 + }, + { + "epoch": 1.0262979196246682, + "grad_norm": 0.4003046452999115, + "learning_rate": 4.5187596729002616e-05, + "loss": 0.0417, + "step": 33250 + }, + { + "epoch": 1.0266065806531268, + "grad_norm": 0.29089632630348206, + "learning_rate": 4.516016757162693e-05, + "loss": 0.0422, + "step": 33260 + }, + { + "epoch": 1.0269152416815852, + "grad_norm": 0.28546735644340515, + "learning_rate": 4.513273988446457e-05, + "loss": 0.0435, + "step": 33270 + }, + { + "epoch": 1.0272239027100438, + "grad_norm": 0.2959470748901367, + "learning_rate": 4.5105313675847296e-05, + "loss": 0.0375, + "step": 33280 + }, + { + "epoch": 1.0275325637385024, + "grad_norm": 0.35182833671569824, + "learning_rate": 4.5077888954106495e-05, + "loss": 0.0429, + "step": 33290 + }, + { + "epoch": 1.0278412247669608, + "grad_norm": 0.3233323097229004, + "learning_rate": 4.505046572757309e-05, + "loss": 0.0409, + "step": 33300 + }, + { + "epoch": 1.0281498857954194, + "grad_norm": 0.4511515200138092, + "learning_rate": 4.502304400457749e-05, + "loss": 0.0439, + "step": 33310 + }, + { + "epoch": 1.028458546823878, + "grad_norm": 0.9808740615844727, + "learning_rate": 4.499562379344973e-05, + "loss": 0.0428, + "step": 33320 + }, + { + "epoch": 1.0287672078523367, + "grad_norm": 0.30109575390815735, + "learning_rate": 4.4968205102519306e-05, + "loss": 0.0399, + "step": 33330 + }, + { + "epoch": 1.029075868880795, + "grad_norm": 0.2891806662082672, + "learning_rate": 4.494078794011532e-05, + "loss": 0.0402, + "step": 33340 + }, + { + "epoch": 1.0293845299092537, + "grad_norm": 0.287392258644104, + "learning_rate": 4.491337231456639e-05, + "loss": 0.0478, + "step": 33350 + }, + { + "epoch": 1.0296931909377123, + "grad_norm": 0.38974466919898987, + "learning_rate": 4.4885958234200634e-05, + "loss": 0.0363, + "step": 33360 + }, + { + "epoch": 1.0300018519661707, + "grad_norm": 0.357264906167984, + "learning_rate": 4.485854570734575e-05, + "loss": 0.0399, + "step": 33370 + }, + { + "epoch": 1.0303105129946293, + "grad_norm": 0.25159695744514465, + "learning_rate": 4.483113474232891e-05, + "loss": 0.0399, + "step": 33380 + }, + { + "epoch": 1.030619174023088, + "grad_norm": 0.26943257451057434, + "learning_rate": 4.480372534747688e-05, + "loss": 0.0363, + "step": 33390 + }, + { + "epoch": 1.0309278350515463, + "grad_norm": 0.4366040527820587, + "learning_rate": 4.477631753111588e-05, + "loss": 0.0388, + "step": 33400 + }, + { + "epoch": 1.031236496080005, + "grad_norm": 0.28816744685173035, + "learning_rate": 4.4748911301571686e-05, + "loss": 0.0436, + "step": 33410 + }, + { + "epoch": 1.0315451571084635, + "grad_norm": 0.3708633780479431, + "learning_rate": 4.472150666716961e-05, + "loss": 0.0426, + "step": 33420 + }, + { + "epoch": 1.0318538181369221, + "grad_norm": 0.3855706453323364, + "learning_rate": 4.469410363623442e-05, + "loss": 0.041, + "step": 33430 + }, + { + "epoch": 1.0321624791653805, + "grad_norm": 0.4038377106189728, + "learning_rate": 4.466670221709044e-05, + "loss": 0.0455, + "step": 33440 + }, + { + "epoch": 1.0324711401938391, + "grad_norm": 0.3428930640220642, + "learning_rate": 4.463930241806154e-05, + "loss": 0.0453, + "step": 33450 + }, + { + "epoch": 1.0327798012222977, + "grad_norm": 0.4365563988685608, + "learning_rate": 4.4611904247471006e-05, + "loss": 0.0407, + "step": 33460 + }, + { + "epoch": 1.0330884622507561, + "grad_norm": 0.4903591275215149, + "learning_rate": 4.458450771364171e-05, + "loss": 0.0409, + "step": 33470 + }, + { + "epoch": 1.0333971232792147, + "grad_norm": 0.31171268224716187, + "learning_rate": 4.4557112824895965e-05, + "loss": 0.0439, + "step": 33480 + }, + { + "epoch": 1.0337057843076733, + "grad_norm": 0.5141140222549438, + "learning_rate": 4.452971958955563e-05, + "loss": 0.0423, + "step": 33490 + }, + { + "epoch": 1.034014445336132, + "grad_norm": 0.2503029406070709, + "learning_rate": 4.450232801594208e-05, + "loss": 0.0424, + "step": 33500 + }, + { + "epoch": 1.0343231063645904, + "grad_norm": 0.28079068660736084, + "learning_rate": 4.447493811237609e-05, + "loss": 0.0423, + "step": 33510 + }, + { + "epoch": 1.034631767393049, + "grad_norm": 0.3369140923023224, + "learning_rate": 4.444754988717804e-05, + "loss": 0.0392, + "step": 33520 + }, + { + "epoch": 1.0349404284215076, + "grad_norm": 0.33987686038017273, + "learning_rate": 4.442016334866771e-05, + "loss": 0.0444, + "step": 33530 + }, + { + "epoch": 1.035249089449966, + "grad_norm": 0.3793368637561798, + "learning_rate": 4.4392778505164445e-05, + "loss": 0.0409, + "step": 33540 + }, + { + "epoch": 1.0355577504784246, + "grad_norm": 0.48139455914497375, + "learning_rate": 4.436539536498702e-05, + "loss": 0.0418, + "step": 33550 + }, + { + "epoch": 1.0358664115068832, + "grad_norm": 0.4044865071773529, + "learning_rate": 4.433801393645369e-05, + "loss": 0.0377, + "step": 33560 + }, + { + "epoch": 1.0361750725353418, + "grad_norm": 0.37945497035980225, + "learning_rate": 4.431063422788226e-05, + "loss": 0.0404, + "step": 33570 + }, + { + "epoch": 1.0364837335638002, + "grad_norm": 0.409257709980011, + "learning_rate": 4.428325624758991e-05, + "loss": 0.0419, + "step": 33580 + }, + { + "epoch": 1.0367923945922588, + "grad_norm": 0.40771231055259705, + "learning_rate": 4.4255880003893366e-05, + "loss": 0.0394, + "step": 33590 + }, + { + "epoch": 1.0371010556207174, + "grad_norm": 0.29461976885795593, + "learning_rate": 4.422850550510884e-05, + "loss": 0.0376, + "step": 33600 + }, + { + "epoch": 1.0374097166491758, + "grad_norm": 0.3269783854484558, + "learning_rate": 4.4201132759551934e-05, + "loss": 0.037, + "step": 33610 + }, + { + "epoch": 1.0377183776776344, + "grad_norm": 0.4535234868526459, + "learning_rate": 4.4173761775537804e-05, + "loss": 0.0402, + "step": 33620 + }, + { + "epoch": 1.038027038706093, + "grad_norm": 0.35693347454071045, + "learning_rate": 4.414639256138099e-05, + "loss": 0.0451, + "step": 33630 + }, + { + "epoch": 1.0383356997345514, + "grad_norm": 0.31567224860191345, + "learning_rate": 4.411902512539557e-05, + "loss": 0.0402, + "step": 33640 + }, + { + "epoch": 1.03864436076301, + "grad_norm": 0.3797275722026825, + "learning_rate": 4.4091659475895044e-05, + "loss": 0.0422, + "step": 33650 + }, + { + "epoch": 1.0389530217914686, + "grad_norm": 0.4938895106315613, + "learning_rate": 4.406429562119235e-05, + "loss": 0.0382, + "step": 33660 + }, + { + "epoch": 1.0392616828199273, + "grad_norm": 0.6031423211097717, + "learning_rate": 4.4036933569599945e-05, + "loss": 0.0424, + "step": 33670 + }, + { + "epoch": 1.0395703438483856, + "grad_norm": 0.2946793735027313, + "learning_rate": 4.400957332942965e-05, + "loss": 0.0392, + "step": 33680 + }, + { + "epoch": 1.0398790048768443, + "grad_norm": 0.2858178913593292, + "learning_rate": 4.3982214908992844e-05, + "loss": 0.0401, + "step": 33690 + }, + { + "epoch": 1.0401876659053029, + "grad_norm": 0.4116646647453308, + "learning_rate": 4.3954858316600235e-05, + "loss": 0.0465, + "step": 33700 + }, + { + "epoch": 1.0404963269337613, + "grad_norm": 0.2933104932308197, + "learning_rate": 4.392750356056205e-05, + "loss": 0.0404, + "step": 33710 + }, + { + "epoch": 1.0408049879622199, + "grad_norm": 0.42587029933929443, + "learning_rate": 4.390015064918798e-05, + "loss": 0.0423, + "step": 33720 + }, + { + "epoch": 1.0411136489906785, + "grad_norm": 0.6634638905525208, + "learning_rate": 4.387279959078705e-05, + "loss": 0.0393, + "step": 33730 + }, + { + "epoch": 1.0414223100191369, + "grad_norm": 0.25277426838874817, + "learning_rate": 4.384545039366786e-05, + "loss": 0.0414, + "step": 33740 + }, + { + "epoch": 1.0417309710475955, + "grad_norm": 0.2534354329109192, + "learning_rate": 4.381810306613831e-05, + "loss": 0.0385, + "step": 33750 + }, + { + "epoch": 1.042039632076054, + "grad_norm": 0.3826468288898468, + "learning_rate": 4.3790757616505826e-05, + "loss": 0.042, + "step": 33760 + }, + { + "epoch": 1.0423482931045127, + "grad_norm": 0.4210922122001648, + "learning_rate": 4.376341405307725e-05, + "loss": 0.0375, + "step": 33770 + }, + { + "epoch": 1.042656954132971, + "grad_norm": 0.3787771761417389, + "learning_rate": 4.37360723841588e-05, + "loss": 0.0453, + "step": 33780 + }, + { + "epoch": 1.0429656151614297, + "grad_norm": 0.48236310482025146, + "learning_rate": 4.370873261805619e-05, + "loss": 0.0409, + "step": 33790 + }, + { + "epoch": 1.0432742761898883, + "grad_norm": 0.34056633710861206, + "learning_rate": 4.368139476307449e-05, + "loss": 0.0425, + "step": 33800 + }, + { + "epoch": 1.0435829372183467, + "grad_norm": 0.3339214026927948, + "learning_rate": 4.365405882751822e-05, + "loss": 0.0384, + "step": 33810 + }, + { + "epoch": 1.0438915982468053, + "grad_norm": 0.4419754147529602, + "learning_rate": 4.3626724819691326e-05, + "loss": 0.0391, + "step": 33820 + }, + { + "epoch": 1.044200259275264, + "grad_norm": 0.4221939146518707, + "learning_rate": 4.359939274789715e-05, + "loss": 0.0422, + "step": 33830 + }, + { + "epoch": 1.0445089203037226, + "grad_norm": 0.29953131079673767, + "learning_rate": 4.357206262043848e-05, + "loss": 0.0386, + "step": 33840 + }, + { + "epoch": 1.044817581332181, + "grad_norm": 0.3198293149471283, + "learning_rate": 4.354473444561745e-05, + "loss": 0.0393, + "step": 33850 + }, + { + "epoch": 1.0451262423606396, + "grad_norm": 0.6314359903335571, + "learning_rate": 4.3517408231735644e-05, + "loss": 0.0421, + "step": 33860 + }, + { + "epoch": 1.0454349033890982, + "grad_norm": 0.2798313796520233, + "learning_rate": 4.3490083987094086e-05, + "loss": 0.0454, + "step": 33870 + }, + { + "epoch": 1.0457435644175566, + "grad_norm": 0.7230936288833618, + "learning_rate": 4.34627617199931e-05, + "loss": 0.0413, + "step": 33880 + }, + { + "epoch": 1.0460522254460152, + "grad_norm": 0.434063583612442, + "learning_rate": 4.3435441438732526e-05, + "loss": 0.0448, + "step": 33890 + }, + { + "epoch": 1.0463608864744738, + "grad_norm": 0.33370983600616455, + "learning_rate": 4.340812315161149e-05, + "loss": 0.0433, + "step": 33900 + }, + { + "epoch": 1.0466695475029322, + "grad_norm": 0.336823433637619, + "learning_rate": 4.338080686692859e-05, + "loss": 0.0397, + "step": 33910 + }, + { + "epoch": 1.0469782085313908, + "grad_norm": 0.37061071395874023, + "learning_rate": 4.3353492592981816e-05, + "loss": 0.0383, + "step": 33920 + }, + { + "epoch": 1.0472868695598494, + "grad_norm": 0.2474900335073471, + "learning_rate": 4.3326180338068485e-05, + "loss": 0.0391, + "step": 33930 + }, + { + "epoch": 1.047595530588308, + "grad_norm": 0.33489155769348145, + "learning_rate": 4.3298870110485356e-05, + "loss": 0.0439, + "step": 33940 + }, + { + "epoch": 1.0479041916167664, + "grad_norm": 0.3254639506340027, + "learning_rate": 4.3271561918528567e-05, + "loss": 0.0412, + "step": 33950 + }, + { + "epoch": 1.048212852645225, + "grad_norm": 0.2773863673210144, + "learning_rate": 4.324425577049359e-05, + "loss": 0.0422, + "step": 33960 + }, + { + "epoch": 1.0485215136736836, + "grad_norm": 0.3373788297176361, + "learning_rate": 4.321695167467535e-05, + "loss": 0.0389, + "step": 33970 + }, + { + "epoch": 1.048830174702142, + "grad_norm": 0.5116084814071655, + "learning_rate": 4.3189649639368093e-05, + "loss": 0.0381, + "step": 33980 + }, + { + "epoch": 1.0491388357306006, + "grad_norm": 0.5037438273429871, + "learning_rate": 4.316234967286547e-05, + "loss": 0.0384, + "step": 33990 + }, + { + "epoch": 1.0494474967590592, + "grad_norm": 0.3752621114253998, + "learning_rate": 4.313505178346046e-05, + "loss": 0.0404, + "step": 34000 + }, + { + "epoch": 1.0497561577875179, + "grad_norm": 0.48332637548446655, + "learning_rate": 4.3107755979445465e-05, + "loss": 0.0425, + "step": 34010 + }, + { + "epoch": 1.0500648188159762, + "grad_norm": 0.3679785132408142, + "learning_rate": 4.308046226911224e-05, + "loss": 0.0391, + "step": 34020 + }, + { + "epoch": 1.0503734798444349, + "grad_norm": 0.3345279097557068, + "learning_rate": 4.305317066075185e-05, + "loss": 0.0444, + "step": 34030 + }, + { + "epoch": 1.0506821408728935, + "grad_norm": 0.32308629155158997, + "learning_rate": 4.302588116265482e-05, + "loss": 0.0397, + "step": 34040 + }, + { + "epoch": 1.0509908019013519, + "grad_norm": 0.47831228375434875, + "learning_rate": 4.299859378311094e-05, + "loss": 0.0456, + "step": 34050 + }, + { + "epoch": 1.0512994629298105, + "grad_norm": 0.30310600996017456, + "learning_rate": 4.2971308530409424e-05, + "loss": 0.0394, + "step": 34060 + }, + { + "epoch": 1.051608123958269, + "grad_norm": 1.9340293407440186, + "learning_rate": 4.2944025412838765e-05, + "loss": 0.0372, + "step": 34070 + }, + { + "epoch": 1.0519167849867275, + "grad_norm": 0.2814205586910248, + "learning_rate": 4.291674443868689e-05, + "loss": 0.0405, + "step": 34080 + }, + { + "epoch": 1.052225446015186, + "grad_norm": 0.3866561949253082, + "learning_rate": 4.288946561624104e-05, + "loss": 0.0433, + "step": 34090 + }, + { + "epoch": 1.0525341070436447, + "grad_norm": 0.24825118482112885, + "learning_rate": 4.2862188953787794e-05, + "loss": 0.0403, + "step": 34100 + }, + { + "epoch": 1.0528427680721033, + "grad_norm": 0.3157804608345032, + "learning_rate": 4.283491445961308e-05, + "loss": 0.0389, + "step": 34110 + }, + { + "epoch": 1.0531514291005617, + "grad_norm": 0.31684330105781555, + "learning_rate": 4.2807642142002155e-05, + "loss": 0.0374, + "step": 34120 + }, + { + "epoch": 1.0534600901290203, + "grad_norm": 0.3012367784976959, + "learning_rate": 4.278037200923966e-05, + "loss": 0.0422, + "step": 34130 + }, + { + "epoch": 1.053768751157479, + "grad_norm": 0.34502145648002625, + "learning_rate": 4.275310406960953e-05, + "loss": 0.0473, + "step": 34140 + }, + { + "epoch": 1.0540774121859373, + "grad_norm": 0.36595359444618225, + "learning_rate": 4.272583833139502e-05, + "loss": 0.0405, + "step": 34150 + }, + { + "epoch": 1.054386073214396, + "grad_norm": 0.36117425560951233, + "learning_rate": 4.2698574802878794e-05, + "loss": 0.0395, + "step": 34160 + }, + { + "epoch": 1.0546947342428545, + "grad_norm": 0.455485463142395, + "learning_rate": 4.2671313492342734e-05, + "loss": 0.0398, + "step": 34170 + }, + { + "epoch": 1.0550033952713131, + "grad_norm": 0.3268142342567444, + "learning_rate": 4.264405440806813e-05, + "loss": 0.0397, + "step": 34180 + }, + { + "epoch": 1.0553120562997715, + "grad_norm": 0.335672527551651, + "learning_rate": 4.26167975583356e-05, + "loss": 0.0411, + "step": 34190 + }, + { + "epoch": 1.0556207173282302, + "grad_norm": 0.3554766774177551, + "learning_rate": 4.2589542951425e-05, + "loss": 0.041, + "step": 34200 + }, + { + "epoch": 1.0559293783566888, + "grad_norm": 0.3020007014274597, + "learning_rate": 4.2562290595615615e-05, + "loss": 0.0444, + "step": 34210 + }, + { + "epoch": 1.0562380393851472, + "grad_norm": 0.38875874876976013, + "learning_rate": 4.2535040499185946e-05, + "loss": 0.0374, + "step": 34220 + }, + { + "epoch": 1.0565467004136058, + "grad_norm": 0.3732885718345642, + "learning_rate": 4.250779267041387e-05, + "loss": 0.0409, + "step": 34230 + }, + { + "epoch": 1.0568553614420644, + "grad_norm": 0.31398946046829224, + "learning_rate": 4.248054711757657e-05, + "loss": 0.0386, + "step": 34240 + }, + { + "epoch": 1.0571640224705228, + "grad_norm": 0.32569220662117004, + "learning_rate": 4.245330384895052e-05, + "loss": 0.0424, + "step": 34250 + }, + { + "epoch": 1.0574726834989814, + "grad_norm": 0.2977937161922455, + "learning_rate": 4.242606287281151e-05, + "loss": 0.0418, + "step": 34260 + }, + { + "epoch": 1.05778134452744, + "grad_norm": 0.3152710199356079, + "learning_rate": 4.2398824197434595e-05, + "loss": 0.0398, + "step": 34270 + }, + { + "epoch": 1.0580900055558986, + "grad_norm": 0.3401305079460144, + "learning_rate": 4.23715878310942e-05, + "loss": 0.0444, + "step": 34280 + }, + { + "epoch": 1.058398666584357, + "grad_norm": 0.35917750000953674, + "learning_rate": 4.234435378206402e-05, + "loss": 0.0397, + "step": 34290 + }, + { + "epoch": 1.0587073276128156, + "grad_norm": 1.1783124208450317, + "learning_rate": 4.2317122058617006e-05, + "loss": 0.0393, + "step": 34300 + }, + { + "epoch": 1.0590159886412742, + "grad_norm": 0.35903316736221313, + "learning_rate": 4.2289892669025485e-05, + "loss": 0.0447, + "step": 34310 + }, + { + "epoch": 1.0593246496697326, + "grad_norm": 0.3458404242992401, + "learning_rate": 4.226266562156097e-05, + "loss": 0.0405, + "step": 34320 + }, + { + "epoch": 1.0596333106981912, + "grad_norm": 0.310506135225296, + "learning_rate": 4.223544092449435e-05, + "loss": 0.045, + "step": 34330 + }, + { + "epoch": 1.0599419717266498, + "grad_norm": 0.28051358461380005, + "learning_rate": 4.2208218586095784e-05, + "loss": 0.0466, + "step": 34340 + }, + { + "epoch": 1.0602506327551084, + "grad_norm": 0.3760857582092285, + "learning_rate": 4.218099861463466e-05, + "loss": 0.0438, + "step": 34350 + }, + { + "epoch": 1.0605592937835668, + "grad_norm": 0.3963478207588196, + "learning_rate": 4.215378101837972e-05, + "loss": 0.0407, + "step": 34360 + }, + { + "epoch": 1.0608679548120254, + "grad_norm": 0.38575127720832825, + "learning_rate": 4.2126565805598937e-05, + "loss": 0.0421, + "step": 34370 + }, + { + "epoch": 1.061176615840484, + "grad_norm": 0.3328394293785095, + "learning_rate": 4.209935298455957e-05, + "loss": 0.04, + "step": 34380 + }, + { + "epoch": 1.0614852768689425, + "grad_norm": 0.44604820013046265, + "learning_rate": 4.207214256352817e-05, + "loss": 0.0416, + "step": 34390 + }, + { + "epoch": 1.061793937897401, + "grad_norm": 0.3029639720916748, + "learning_rate": 4.2044934550770524e-05, + "loss": 0.0429, + "step": 34400 + }, + { + "epoch": 1.0621025989258597, + "grad_norm": 0.46753761172294617, + "learning_rate": 4.201772895455174e-05, + "loss": 0.0474, + "step": 34410 + }, + { + "epoch": 1.062411259954318, + "grad_norm": 0.25181010365486145, + "learning_rate": 4.199052578313613e-05, + "loss": 0.0381, + "step": 34420 + }, + { + "epoch": 1.0627199209827767, + "grad_norm": 0.4159124791622162, + "learning_rate": 4.1963325044787294e-05, + "loss": 0.0355, + "step": 34430 + }, + { + "epoch": 1.0630285820112353, + "grad_norm": 0.21723636984825134, + "learning_rate": 4.193612674776814e-05, + "loss": 0.0407, + "step": 34440 + }, + { + "epoch": 1.063337243039694, + "grad_norm": 0.3602229654788971, + "learning_rate": 4.1908930900340745e-05, + "loss": 0.0415, + "step": 34450 + }, + { + "epoch": 1.0636459040681523, + "grad_norm": 0.3095686733722687, + "learning_rate": 4.1881737510766536e-05, + "loss": 0.0412, + "step": 34460 + }, + { + "epoch": 1.063954565096611, + "grad_norm": 0.39142003655433655, + "learning_rate": 4.185454658730609e-05, + "loss": 0.0403, + "step": 34470 + }, + { + "epoch": 1.0642632261250695, + "grad_norm": 1.2974786758422852, + "learning_rate": 4.1827358138219355e-05, + "loss": 0.0424, + "step": 34480 + }, + { + "epoch": 1.064571887153528, + "grad_norm": 0.2877868115901947, + "learning_rate": 4.1800172171765404e-05, + "loss": 0.0401, + "step": 34490 + }, + { + "epoch": 1.0648805481819865, + "grad_norm": 0.3992677628993988, + "learning_rate": 4.177298869620264e-05, + "loss": 0.0449, + "step": 34500 + }, + { + "epoch": 1.0651892092104451, + "grad_norm": 0.46115627884864807, + "learning_rate": 4.1745807719788705e-05, + "loss": 0.042, + "step": 34510 + }, + { + "epoch": 1.0654978702389037, + "grad_norm": 0.3663972020149231, + "learning_rate": 4.1718629250780445e-05, + "loss": 0.0385, + "step": 34520 + }, + { + "epoch": 1.0658065312673621, + "grad_norm": 0.25777730345726013, + "learning_rate": 4.1691453297433956e-05, + "loss": 0.0413, + "step": 34530 + }, + { + "epoch": 1.0661151922958207, + "grad_norm": 0.33440643548965454, + "learning_rate": 4.166427986800457e-05, + "loss": 0.0446, + "step": 34540 + }, + { + "epoch": 1.0664238533242794, + "grad_norm": 0.5503690838813782, + "learning_rate": 4.163710897074688e-05, + "loss": 0.0417, + "step": 34550 + }, + { + "epoch": 1.0667325143527377, + "grad_norm": 0.35317155718803406, + "learning_rate": 4.1609940613914686e-05, + "loss": 0.042, + "step": 34560 + }, + { + "epoch": 1.0670411753811964, + "grad_norm": 0.34216102957725525, + "learning_rate": 4.1582774805760996e-05, + "loss": 0.0392, + "step": 34570 + }, + { + "epoch": 1.067349836409655, + "grad_norm": 0.35230669379234314, + "learning_rate": 4.155561155453809e-05, + "loss": 0.0417, + "step": 34580 + }, + { + "epoch": 1.0676584974381134, + "grad_norm": 0.29203304648399353, + "learning_rate": 4.15284508684974e-05, + "loss": 0.0446, + "step": 34590 + }, + { + "epoch": 1.067967158466572, + "grad_norm": 0.4045337736606598, + "learning_rate": 4.1501292755889675e-05, + "loss": 0.0393, + "step": 34600 + }, + { + "epoch": 1.0682758194950306, + "grad_norm": 0.3442783057689667, + "learning_rate": 4.1474137224964833e-05, + "loss": 0.0406, + "step": 34610 + }, + { + "epoch": 1.0685844805234892, + "grad_norm": 3.7407217025756836, + "learning_rate": 4.144698428397197e-05, + "loss": 0.0398, + "step": 34620 + }, + { + "epoch": 1.0688931415519476, + "grad_norm": 0.277869313955307, + "learning_rate": 4.1419833941159466e-05, + "loss": 0.0391, + "step": 34630 + }, + { + "epoch": 1.0692018025804062, + "grad_norm": 0.3726458251476288, + "learning_rate": 4.1392686204774846e-05, + "loss": 0.0423, + "step": 34640 + }, + { + "epoch": 1.0695104636088648, + "grad_norm": 0.32113856077194214, + "learning_rate": 4.13655410830649e-05, + "loss": 0.0437, + "step": 34650 + }, + { + "epoch": 1.0698191246373232, + "grad_norm": 0.3391062915325165, + "learning_rate": 4.1338398584275594e-05, + "loss": 0.0379, + "step": 34660 + }, + { + "epoch": 1.0701277856657818, + "grad_norm": 0.28679537773132324, + "learning_rate": 4.1311258716652104e-05, + "loss": 0.0401, + "step": 34670 + }, + { + "epoch": 1.0704364466942404, + "grad_norm": 0.32151272892951965, + "learning_rate": 4.128412148843881e-05, + "loss": 0.0393, + "step": 34680 + }, + { + "epoch": 1.0707451077226988, + "grad_norm": 0.33013856410980225, + "learning_rate": 4.125698690787926e-05, + "loss": 0.0395, + "step": 34690 + }, + { + "epoch": 1.0710537687511574, + "grad_norm": 0.3058496117591858, + "learning_rate": 4.1229854983216245e-05, + "loss": 0.0397, + "step": 34700 + }, + { + "epoch": 1.071362429779616, + "grad_norm": 0.32587623596191406, + "learning_rate": 4.120272572269175e-05, + "loss": 0.04, + "step": 34710 + }, + { + "epoch": 1.0716710908080747, + "grad_norm": 0.3043861389160156, + "learning_rate": 4.117559913454687e-05, + "loss": 0.0399, + "step": 34720 + }, + { + "epoch": 1.071979751836533, + "grad_norm": 0.35377731919288635, + "learning_rate": 4.114847522702201e-05, + "loss": 0.0403, + "step": 34730 + }, + { + "epoch": 1.0722884128649917, + "grad_norm": 0.3600635230541229, + "learning_rate": 4.112135400835664e-05, + "loss": 0.0447, + "step": 34740 + }, + { + "epoch": 1.0725970738934503, + "grad_norm": 0.3468264043331146, + "learning_rate": 4.109423548678949e-05, + "loss": 0.0426, + "step": 34750 + }, + { + "epoch": 1.0729057349219087, + "grad_norm": 0.3154706656932831, + "learning_rate": 4.106711967055848e-05, + "loss": 0.0462, + "step": 34760 + }, + { + "epoch": 1.0732143959503673, + "grad_norm": 0.36817818880081177, + "learning_rate": 4.1040006567900636e-05, + "loss": 0.0394, + "step": 34770 + }, + { + "epoch": 1.0735230569788259, + "grad_norm": 0.34683725237846375, + "learning_rate": 4.101289618705224e-05, + "loss": 0.0421, + "step": 34780 + }, + { + "epoch": 1.0738317180072845, + "grad_norm": 0.3008275330066681, + "learning_rate": 4.0985788536248675e-05, + "loss": 0.0427, + "step": 34790 + }, + { + "epoch": 1.0741403790357429, + "grad_norm": 0.30643412470817566, + "learning_rate": 4.095868362372454e-05, + "loss": 0.0402, + "step": 34800 + }, + { + "epoch": 1.0744490400642015, + "grad_norm": 0.34412944316864014, + "learning_rate": 4.0931581457713614e-05, + "loss": 0.0436, + "step": 34810 + }, + { + "epoch": 1.0747577010926601, + "grad_norm": 0.27555081248283386, + "learning_rate": 4.09044820464488e-05, + "loss": 0.042, + "step": 34820 + }, + { + "epoch": 1.0750663621211185, + "grad_norm": 0.3463650643825531, + "learning_rate": 4.087738539816219e-05, + "loss": 0.0387, + "step": 34830 + }, + { + "epoch": 1.0753750231495771, + "grad_norm": 0.3320775032043457, + "learning_rate": 4.085029152108501e-05, + "loss": 0.0372, + "step": 34840 + }, + { + "epoch": 1.0756836841780357, + "grad_norm": 0.45284217596054077, + "learning_rate": 4.0823200423447714e-05, + "loss": 0.0389, + "step": 34850 + }, + { + "epoch": 1.0759923452064943, + "grad_norm": 0.39344990253448486, + "learning_rate": 4.079611211347981e-05, + "loss": 0.0401, + "step": 34860 + }, + { + "epoch": 1.0763010062349527, + "grad_norm": 0.41826385259628296, + "learning_rate": 4.076902659941002e-05, + "loss": 0.0368, + "step": 34870 + }, + { + "epoch": 1.0766096672634113, + "grad_norm": 0.3596637547016144, + "learning_rate": 4.074194388946624e-05, + "loss": 0.0423, + "step": 34880 + }, + { + "epoch": 1.07691832829187, + "grad_norm": 0.3408035337924957, + "learning_rate": 4.071486399187545e-05, + "loss": 0.0385, + "step": 34890 + }, + { + "epoch": 1.0772269893203283, + "grad_norm": 0.3208141624927521, + "learning_rate": 4.0687786914863836e-05, + "loss": 0.0414, + "step": 34900 + }, + { + "epoch": 1.077535650348787, + "grad_norm": 0.2877424359321594, + "learning_rate": 4.0660712666656666e-05, + "loss": 0.0402, + "step": 34910 + }, + { + "epoch": 1.0778443113772456, + "grad_norm": 0.2790810465812683, + "learning_rate": 4.0633641255478394e-05, + "loss": 0.0383, + "step": 34920 + }, + { + "epoch": 1.078152972405704, + "grad_norm": 0.3657786250114441, + "learning_rate": 4.0606572689552624e-05, + "loss": 0.0411, + "step": 34930 + }, + { + "epoch": 1.0784616334341626, + "grad_norm": 0.3321923017501831, + "learning_rate": 4.0579506977102036e-05, + "loss": 0.0436, + "step": 34940 + }, + { + "epoch": 1.0787702944626212, + "grad_norm": 0.42663681507110596, + "learning_rate": 4.055244412634849e-05, + "loss": 0.0423, + "step": 34950 + }, + { + "epoch": 1.0790789554910798, + "grad_norm": 0.774666965007782, + "learning_rate": 4.052538414551298e-05, + "loss": 0.0375, + "step": 34960 + }, + { + "epoch": 1.0793876165195382, + "grad_norm": 0.3703368604183197, + "learning_rate": 4.0498327042815596e-05, + "loss": 0.0438, + "step": 34970 + }, + { + "epoch": 1.0796962775479968, + "grad_norm": 0.39216116070747375, + "learning_rate": 4.047127282647559e-05, + "loss": 0.0379, + "step": 34980 + }, + { + "epoch": 1.0800049385764554, + "grad_norm": 0.32482266426086426, + "learning_rate": 4.04442215047113e-05, + "loss": 0.0421, + "step": 34990 + }, + { + "epoch": 1.0803135996049138, + "grad_norm": 0.3227287828922272, + "learning_rate": 4.041717308574023e-05, + "loss": 0.0442, + "step": 35000 + }, + { + "epoch": 1.0806222606333724, + "grad_norm": 0.3611975312232971, + "learning_rate": 4.039012757777893e-05, + "loss": 0.035, + "step": 35010 + }, + { + "epoch": 1.080930921661831, + "grad_norm": 0.33381038904190063, + "learning_rate": 4.036308498904314e-05, + "loss": 0.0403, + "step": 35020 + }, + { + "epoch": 1.0812395826902894, + "grad_norm": 0.30908524990081787, + "learning_rate": 4.033604532774771e-05, + "loss": 0.0404, + "step": 35030 + }, + { + "epoch": 1.081548243718748, + "grad_norm": 0.3083938658237457, + "learning_rate": 4.030900860210652e-05, + "loss": 0.0346, + "step": 35040 + }, + { + "epoch": 1.0818569047472066, + "grad_norm": 1.1964077949523926, + "learning_rate": 4.028197482033266e-05, + "loss": 0.0414, + "step": 35050 + }, + { + "epoch": 1.0821655657756653, + "grad_norm": 0.3921232223510742, + "learning_rate": 4.0254943990638246e-05, + "loss": 0.0404, + "step": 35060 + }, + { + "epoch": 1.0824742268041236, + "grad_norm": 0.4027557373046875, + "learning_rate": 4.022791612123454e-05, + "loss": 0.0438, + "step": 35070 + }, + { + "epoch": 1.0827828878325823, + "grad_norm": 0.2793789207935333, + "learning_rate": 4.020089122033192e-05, + "loss": 0.0417, + "step": 35080 + }, + { + "epoch": 1.0830915488610409, + "grad_norm": 0.33566856384277344, + "learning_rate": 4.01738692961398e-05, + "loss": 0.0389, + "step": 35090 + }, + { + "epoch": 1.0834002098894993, + "grad_norm": 0.3540512025356293, + "learning_rate": 4.014685035686675e-05, + "loss": 0.0395, + "step": 35100 + }, + { + "epoch": 1.0837088709179579, + "grad_norm": 0.48548027873039246, + "learning_rate": 4.011983441072039e-05, + "loss": 0.0418, + "step": 35110 + }, + { + "epoch": 1.0840175319464165, + "grad_norm": 0.5770190954208374, + "learning_rate": 4.0092821465907485e-05, + "loss": 0.0397, + "step": 35120 + }, + { + "epoch": 1.084326192974875, + "grad_norm": 0.38621947169303894, + "learning_rate": 4.006581153063383e-05, + "loss": 0.0374, + "step": 35130 + }, + { + "epoch": 1.0846348540033335, + "grad_norm": 0.4722742438316345, + "learning_rate": 4.003880461310432e-05, + "loss": 0.0361, + "step": 35140 + }, + { + "epoch": 1.084943515031792, + "grad_norm": 0.7207523584365845, + "learning_rate": 4.001180072152298e-05, + "loss": 0.0375, + "step": 35150 + }, + { + "epoch": 1.0852521760602507, + "grad_norm": 0.33984893560409546, + "learning_rate": 3.998479986409285e-05, + "loss": 0.0399, + "step": 35160 + }, + { + "epoch": 1.085560837088709, + "grad_norm": 0.32592591643333435, + "learning_rate": 3.995780204901607e-05, + "loss": 0.0346, + "step": 35170 + }, + { + "epoch": 1.0858694981171677, + "grad_norm": 0.36210036277770996, + "learning_rate": 3.993080728449391e-05, + "loss": 0.0402, + "step": 35180 + }, + { + "epoch": 1.0861781591456263, + "grad_norm": 0.30389660596847534, + "learning_rate": 3.990381557872661e-05, + "loss": 0.0377, + "step": 35190 + }, + { + "epoch": 1.086486820174085, + "grad_norm": 0.32349884510040283, + "learning_rate": 3.987682693991359e-05, + "loss": 0.0385, + "step": 35200 + }, + { + "epoch": 1.0867954812025433, + "grad_norm": 0.3543013036251068, + "learning_rate": 3.9849841376253226e-05, + "loss": 0.0433, + "step": 35210 + }, + { + "epoch": 1.087104142231002, + "grad_norm": 0.31742411851882935, + "learning_rate": 3.982285889594306e-05, + "loss": 0.0384, + "step": 35220 + }, + { + "epoch": 1.0874128032594605, + "grad_norm": 1.7199143171310425, + "learning_rate": 3.9795879507179665e-05, + "loss": 0.04, + "step": 35230 + }, + { + "epoch": 1.087721464287919, + "grad_norm": 0.4247044622898102, + "learning_rate": 3.9768903218158634e-05, + "loss": 0.0456, + "step": 35240 + }, + { + "epoch": 1.0880301253163775, + "grad_norm": 0.35673075914382935, + "learning_rate": 3.974193003707468e-05, + "loss": 0.0359, + "step": 35250 + }, + { + "epoch": 1.0883387863448362, + "grad_norm": 0.4697110950946808, + "learning_rate": 3.971495997212152e-05, + "loss": 0.0365, + "step": 35260 + }, + { + "epoch": 1.0886474473732946, + "grad_norm": 0.2623203992843628, + "learning_rate": 3.9687993031491985e-05, + "loss": 0.0355, + "step": 35270 + }, + { + "epoch": 1.0889561084017532, + "grad_norm": 0.3344101011753082, + "learning_rate": 3.966102922337787e-05, + "loss": 0.0426, + "step": 35280 + }, + { + "epoch": 1.0892647694302118, + "grad_norm": 0.2960200905799866, + "learning_rate": 3.963406855597009e-05, + "loss": 0.0415, + "step": 35290 + }, + { + "epoch": 1.0895734304586704, + "grad_norm": 0.41818156838417053, + "learning_rate": 3.960711103745861e-05, + "loss": 0.0409, + "step": 35300 + }, + { + "epoch": 1.0898820914871288, + "grad_norm": 1.923719048500061, + "learning_rate": 3.958015667603237e-05, + "loss": 0.0392, + "step": 35310 + }, + { + "epoch": 1.0901907525155874, + "grad_norm": 0.32413050532341003, + "learning_rate": 3.955320547987943e-05, + "loss": 0.0428, + "step": 35320 + }, + { + "epoch": 1.090499413544046, + "grad_norm": 0.5209041833877563, + "learning_rate": 3.952625745718681e-05, + "loss": 0.045, + "step": 35330 + }, + { + "epoch": 1.0908080745725044, + "grad_norm": 0.39176860451698303, + "learning_rate": 3.949931261614064e-05, + "loss": 0.0422, + "step": 35340 + }, + { + "epoch": 1.091116735600963, + "grad_norm": 0.302581787109375, + "learning_rate": 3.947237096492605e-05, + "loss": 0.0409, + "step": 35350 + }, + { + "epoch": 1.0914253966294216, + "grad_norm": 0.42392462491989136, + "learning_rate": 3.944543251172719e-05, + "loss": 0.042, + "step": 35360 + }, + { + "epoch": 1.09173405765788, + "grad_norm": 0.2089165300130844, + "learning_rate": 3.941849726472725e-05, + "loss": 0.0443, + "step": 35370 + }, + { + "epoch": 1.0920427186863386, + "grad_norm": 0.382326602935791, + "learning_rate": 3.939156523210846e-05, + "loss": 0.0397, + "step": 35380 + }, + { + "epoch": 1.0923513797147972, + "grad_norm": 0.3986291289329529, + "learning_rate": 3.9364636422052046e-05, + "loss": 0.0376, + "step": 35390 + }, + { + "epoch": 1.0926600407432558, + "grad_norm": 0.3733333945274353, + "learning_rate": 3.933771084273828e-05, + "loss": 0.0403, + "step": 35400 + }, + { + "epoch": 1.0929687017717142, + "grad_norm": 0.38256412744522095, + "learning_rate": 3.931078850234643e-05, + "loss": 0.0453, + "step": 35410 + }, + { + "epoch": 1.0932773628001728, + "grad_norm": 0.2882813513278961, + "learning_rate": 3.928386940905483e-05, + "loss": 0.0421, + "step": 35420 + }, + { + "epoch": 1.0935860238286315, + "grad_norm": 0.30263057351112366, + "learning_rate": 3.925695357104073e-05, + "loss": 0.0406, + "step": 35430 + }, + { + "epoch": 1.0938946848570898, + "grad_norm": 0.32856374979019165, + "learning_rate": 3.923004099648049e-05, + "loss": 0.04, + "step": 35440 + }, + { + "epoch": 1.0942033458855485, + "grad_norm": 0.3419033885002136, + "learning_rate": 3.920313169354944e-05, + "loss": 0.0361, + "step": 35450 + }, + { + "epoch": 1.094512006914007, + "grad_norm": 0.4027165174484253, + "learning_rate": 3.9176225670421897e-05, + "loss": 0.038, + "step": 35460 + }, + { + "epoch": 1.0948206679424657, + "grad_norm": 0.26680031418800354, + "learning_rate": 3.9149322935271224e-05, + "loss": 0.044, + "step": 35470 + }, + { + "epoch": 1.095129328970924, + "grad_norm": 0.3616836965084076, + "learning_rate": 3.9122423496269725e-05, + "loss": 0.036, + "step": 35480 + }, + { + "epoch": 1.0954379899993827, + "grad_norm": 0.28829699754714966, + "learning_rate": 3.909552736158877e-05, + "loss": 0.0429, + "step": 35490 + }, + { + "epoch": 1.0957466510278413, + "grad_norm": 0.28914961218833923, + "learning_rate": 3.90686345393987e-05, + "loss": 0.0395, + "step": 35500 + }, + { + "epoch": 1.0960553120562997, + "grad_norm": 0.32090476155281067, + "learning_rate": 3.9041745037868816e-05, + "loss": 0.035, + "step": 35510 + }, + { + "epoch": 1.0963639730847583, + "grad_norm": 0.46613141894340515, + "learning_rate": 3.9014858865167465e-05, + "loss": 0.0457, + "step": 35520 + }, + { + "epoch": 1.096672634113217, + "grad_norm": 0.32297348976135254, + "learning_rate": 3.8987976029461935e-05, + "loss": 0.0401, + "step": 35530 + }, + { + "epoch": 1.0969812951416755, + "grad_norm": 0.263986736536026, + "learning_rate": 3.896109653891853e-05, + "loss": 0.0405, + "step": 35540 + }, + { + "epoch": 1.097289956170134, + "grad_norm": 0.3686315417289734, + "learning_rate": 3.893422040170254e-05, + "loss": 0.0392, + "step": 35550 + }, + { + "epoch": 1.0975986171985925, + "grad_norm": 0.2934940755367279, + "learning_rate": 3.8907347625978207e-05, + "loss": 0.0402, + "step": 35560 + }, + { + "epoch": 1.0979072782270511, + "grad_norm": 0.2970115840435028, + "learning_rate": 3.88804782199088e-05, + "loss": 0.0391, + "step": 35570 + }, + { + "epoch": 1.0982159392555095, + "grad_norm": 0.34263888001441956, + "learning_rate": 3.8853612191656495e-05, + "loss": 0.0418, + "step": 35580 + }, + { + "epoch": 1.0985246002839681, + "grad_norm": 0.37717363238334656, + "learning_rate": 3.88267495493825e-05, + "loss": 0.0403, + "step": 35590 + }, + { + "epoch": 1.0988332613124268, + "grad_norm": 0.36386755108833313, + "learning_rate": 3.8799890301247004e-05, + "loss": 0.041, + "step": 35600 + }, + { + "epoch": 1.0991419223408851, + "grad_norm": 0.3352445065975189, + "learning_rate": 3.8773034455409096e-05, + "loss": 0.0416, + "step": 35610 + }, + { + "epoch": 1.0994505833693438, + "grad_norm": 0.338324636220932, + "learning_rate": 3.8746182020026904e-05, + "loss": 0.0388, + "step": 35620 + }, + { + "epoch": 1.0997592443978024, + "grad_norm": 0.3151583969593048, + "learning_rate": 3.871933300325745e-05, + "loss": 0.0392, + "step": 35630 + }, + { + "epoch": 1.100067905426261, + "grad_norm": 0.4615723788738251, + "learning_rate": 3.869248741325679e-05, + "loss": 0.0402, + "step": 35640 + }, + { + "epoch": 1.1003765664547194, + "grad_norm": 0.3377203047275543, + "learning_rate": 3.866564525817992e-05, + "loss": 0.0413, + "step": 35650 + }, + { + "epoch": 1.100685227483178, + "grad_norm": 0.49120479822158813, + "learning_rate": 3.8638806546180725e-05, + "loss": 0.0427, + "step": 35660 + }, + { + "epoch": 1.1009938885116366, + "grad_norm": 0.39829352498054504, + "learning_rate": 3.861197128541213e-05, + "loss": 0.0403, + "step": 35670 + }, + { + "epoch": 1.101302549540095, + "grad_norm": 0.3715193569660187, + "learning_rate": 3.858513948402599e-05, + "loss": 0.0401, + "step": 35680 + }, + { + "epoch": 1.1016112105685536, + "grad_norm": 0.5617470741271973, + "learning_rate": 3.8558311150173077e-05, + "loss": 0.0413, + "step": 35690 + }, + { + "epoch": 1.1019198715970122, + "grad_norm": 0.3019670248031616, + "learning_rate": 3.853148629200312e-05, + "loss": 0.0383, + "step": 35700 + }, + { + "epoch": 1.1022285326254706, + "grad_norm": 0.5189443230628967, + "learning_rate": 3.850466491766482e-05, + "loss": 0.0376, + "step": 35710 + }, + { + "epoch": 1.1025371936539292, + "grad_norm": 0.5236128568649292, + "learning_rate": 3.847784703530583e-05, + "loss": 0.0377, + "step": 35720 + }, + { + "epoch": 1.1028458546823878, + "grad_norm": 0.3059752583503723, + "learning_rate": 3.845103265307266e-05, + "loss": 0.0377, + "step": 35730 + }, + { + "epoch": 1.1031545157108464, + "grad_norm": 0.598954975605011, + "learning_rate": 3.842422177911086e-05, + "loss": 0.0383, + "step": 35740 + }, + { + "epoch": 1.1034631767393048, + "grad_norm": 0.2638494670391083, + "learning_rate": 3.8397414421564826e-05, + "loss": 0.0394, + "step": 35750 + }, + { + "epoch": 1.1037718377677634, + "grad_norm": 0.34799724817276, + "learning_rate": 3.8370610588577935e-05, + "loss": 0.0388, + "step": 35760 + }, + { + "epoch": 1.104080498796222, + "grad_norm": 0.37238264083862305, + "learning_rate": 3.834381028829251e-05, + "loss": 0.0416, + "step": 35770 + }, + { + "epoch": 1.1043891598246804, + "grad_norm": 0.26326724886894226, + "learning_rate": 3.8317013528849745e-05, + "loss": 0.037, + "step": 35780 + }, + { + "epoch": 1.104697820853139, + "grad_norm": 0.272718220949173, + "learning_rate": 3.8290220318389815e-05, + "loss": 0.0382, + "step": 35790 + }, + { + "epoch": 1.1050064818815977, + "grad_norm": 0.33146414160728455, + "learning_rate": 3.8263430665051746e-05, + "loss": 0.0365, + "step": 35800 + }, + { + "epoch": 1.1053151429100563, + "grad_norm": 0.4555759131908417, + "learning_rate": 3.8236644576973554e-05, + "loss": 0.0402, + "step": 35810 + }, + { + "epoch": 1.1056238039385147, + "grad_norm": 0.37569403648376465, + "learning_rate": 3.820986206229217e-05, + "loss": 0.0415, + "step": 35820 + }, + { + "epoch": 1.1059324649669733, + "grad_norm": 0.2865109443664551, + "learning_rate": 3.8183083129143384e-05, + "loss": 0.039, + "step": 35830 + }, + { + "epoch": 1.106241125995432, + "grad_norm": 0.25734609365463257, + "learning_rate": 3.815630778566193e-05, + "loss": 0.0371, + "step": 35840 + }, + { + "epoch": 1.1065497870238903, + "grad_norm": 0.2770327031612396, + "learning_rate": 3.812953603998145e-05, + "loss": 0.0386, + "step": 35850 + }, + { + "epoch": 1.106858448052349, + "grad_norm": 0.3009486794471741, + "learning_rate": 3.8102767900234504e-05, + "loss": 0.0379, + "step": 35860 + }, + { + "epoch": 1.1071671090808075, + "grad_norm": 0.3634641766548157, + "learning_rate": 3.807600337455256e-05, + "loss": 0.041, + "step": 35870 + }, + { + "epoch": 1.107475770109266, + "grad_norm": 0.25677016377449036, + "learning_rate": 3.804924247106593e-05, + "loss": 0.0349, + "step": 35880 + }, + { + "epoch": 1.1077844311377245, + "grad_norm": 0.39879438281059265, + "learning_rate": 3.8022485197903925e-05, + "loss": 0.0393, + "step": 35890 + }, + { + "epoch": 1.1080930921661831, + "grad_norm": 0.4078949987888336, + "learning_rate": 3.799573156319464e-05, + "loss": 0.0403, + "step": 35900 + }, + { + "epoch": 1.1084017531946417, + "grad_norm": 0.43594446778297424, + "learning_rate": 3.796898157506515e-05, + "loss": 0.0385, + "step": 35910 + }, + { + "epoch": 1.1087104142231001, + "grad_norm": 0.3061947524547577, + "learning_rate": 3.794223524164143e-05, + "loss": 0.0397, + "step": 35920 + }, + { + "epoch": 1.1090190752515587, + "grad_norm": 0.31529802083969116, + "learning_rate": 3.7915492571048245e-05, + "loss": 0.0364, + "step": 35930 + }, + { + "epoch": 1.1093277362800174, + "grad_norm": 0.303058922290802, + "learning_rate": 3.788875357140937e-05, + "loss": 0.0404, + "step": 35940 + }, + { + "epoch": 1.1096363973084757, + "grad_norm": 0.3331080377101898, + "learning_rate": 3.786201825084736e-05, + "loss": 0.0395, + "step": 35950 + }, + { + "epoch": 1.1099450583369344, + "grad_norm": 0.5629733204841614, + "learning_rate": 3.783528661748372e-05, + "loss": 0.0349, + "step": 35960 + }, + { + "epoch": 1.110253719365393, + "grad_norm": 0.2780642807483673, + "learning_rate": 3.780855867943882e-05, + "loss": 0.0398, + "step": 35970 + }, + { + "epoch": 1.1105623803938514, + "grad_norm": 0.32601627707481384, + "learning_rate": 3.778183444483189e-05, + "loss": 0.0445, + "step": 35980 + }, + { + "epoch": 1.11087104142231, + "grad_norm": 0.3328104317188263, + "learning_rate": 3.775511392178108e-05, + "loss": 0.0414, + "step": 35990 + }, + { + "epoch": 1.1111797024507686, + "grad_norm": 0.3773289620876312, + "learning_rate": 3.772839711840332e-05, + "loss": 0.0458, + "step": 36000 + }, + { + "epoch": 1.1114883634792272, + "grad_norm": 0.4770675599575043, + "learning_rate": 3.7701684042814515e-05, + "loss": 0.0392, + "step": 36010 + }, + { + "epoch": 1.1117970245076856, + "grad_norm": 0.4023846983909607, + "learning_rate": 3.76749747031294e-05, + "loss": 0.0418, + "step": 36020 + }, + { + "epoch": 1.1121056855361442, + "grad_norm": 0.8341463208198547, + "learning_rate": 3.764826910746152e-05, + "loss": 0.038, + "step": 36030 + }, + { + "epoch": 1.1124143465646028, + "grad_norm": 0.3555240035057068, + "learning_rate": 3.762156726392338e-05, + "loss": 0.0373, + "step": 36040 + }, + { + "epoch": 1.1127230075930612, + "grad_norm": 0.3756515085697174, + "learning_rate": 3.759486918062625e-05, + "loss": 0.0397, + "step": 36050 + }, + { + "epoch": 1.1130316686215198, + "grad_norm": 0.24656088650226593, + "learning_rate": 3.756817486568033e-05, + "loss": 0.0395, + "step": 36060 + }, + { + "epoch": 1.1133403296499784, + "grad_norm": 0.4034716784954071, + "learning_rate": 3.7541484327194654e-05, + "loss": 0.0385, + "step": 36070 + }, + { + "epoch": 1.113648990678437, + "grad_norm": 0.3145918548107147, + "learning_rate": 3.751479757327707e-05, + "loss": 0.0417, + "step": 36080 + }, + { + "epoch": 1.1139576517068954, + "grad_norm": 0.2717500925064087, + "learning_rate": 3.7488114612034345e-05, + "loss": 0.0386, + "step": 36090 + }, + { + "epoch": 1.114266312735354, + "grad_norm": 0.38613060116767883, + "learning_rate": 3.7461435451572044e-05, + "loss": 0.0418, + "step": 36100 + }, + { + "epoch": 1.1145749737638126, + "grad_norm": 0.2729487717151642, + "learning_rate": 3.743476009999459e-05, + "loss": 0.0395, + "step": 36110 + }, + { + "epoch": 1.114883634792271, + "grad_norm": 0.22989381849765778, + "learning_rate": 3.7408088565405245e-05, + "loss": 0.0377, + "step": 36120 + }, + { + "epoch": 1.1151922958207297, + "grad_norm": 0.449336439371109, + "learning_rate": 3.738142085590612e-05, + "loss": 0.0389, + "step": 36130 + }, + { + "epoch": 1.1155009568491883, + "grad_norm": 0.35332852602005005, + "learning_rate": 3.7354756979598194e-05, + "loss": 0.0337, + "step": 36140 + }, + { + "epoch": 1.1158096178776469, + "grad_norm": 0.27218955755233765, + "learning_rate": 3.7328096944581187e-05, + "loss": 0.0396, + "step": 36150 + }, + { + "epoch": 1.1161182789061053, + "grad_norm": 0.43332362174987793, + "learning_rate": 3.730144075895377e-05, + "loss": 0.0392, + "step": 36160 + }, + { + "epoch": 1.1164269399345639, + "grad_norm": 0.2983797490596771, + "learning_rate": 3.727478843081335e-05, + "loss": 0.0432, + "step": 36170 + }, + { + "epoch": 1.1167356009630225, + "grad_norm": 0.2825954258441925, + "learning_rate": 3.72481399682562e-05, + "loss": 0.0413, + "step": 36180 + }, + { + "epoch": 1.1170442619914809, + "grad_norm": 0.9941824078559875, + "learning_rate": 3.722149537937747e-05, + "loss": 0.0395, + "step": 36190 + }, + { + "epoch": 1.1173529230199395, + "grad_norm": 0.3483797609806061, + "learning_rate": 3.7194854672271015e-05, + "loss": 0.0415, + "step": 36200 + }, + { + "epoch": 1.117661584048398, + "grad_norm": 0.29284536838531494, + "learning_rate": 3.7168217855029644e-05, + "loss": 0.0418, + "step": 36210 + }, + { + "epoch": 1.1179702450768565, + "grad_norm": 0.30460554361343384, + "learning_rate": 3.7141584935744856e-05, + "loss": 0.0415, + "step": 36220 + }, + { + "epoch": 1.118278906105315, + "grad_norm": 0.278720885515213, + "learning_rate": 3.7114955922507055e-05, + "loss": 0.042, + "step": 36230 + }, + { + "epoch": 1.1185875671337737, + "grad_norm": 0.31237322092056274, + "learning_rate": 3.708833082340545e-05, + "loss": 0.0429, + "step": 36240 + }, + { + "epoch": 1.1188962281622323, + "grad_norm": 0.4922821521759033, + "learning_rate": 3.7061709646528034e-05, + "loss": 0.0384, + "step": 36250 + }, + { + "epoch": 1.1192048891906907, + "grad_norm": 0.3026488125324249, + "learning_rate": 3.7035092399961604e-05, + "loss": 0.0396, + "step": 36260 + }, + { + "epoch": 1.1195135502191493, + "grad_norm": 0.5608983635902405, + "learning_rate": 3.700847909179177e-05, + "loss": 0.0392, + "step": 36270 + }, + { + "epoch": 1.119822211247608, + "grad_norm": 0.37580540776252747, + "learning_rate": 3.698186973010297e-05, + "loss": 0.042, + "step": 36280 + }, + { + "epoch": 1.1201308722760663, + "grad_norm": 0.3553345799446106, + "learning_rate": 3.695526432297844e-05, + "loss": 0.044, + "step": 36290 + }, + { + "epoch": 1.120439533304525, + "grad_norm": 0.26548218727111816, + "learning_rate": 3.692866287850017e-05, + "loss": 0.038, + "step": 36300 + }, + { + "epoch": 1.1207481943329836, + "grad_norm": 0.2866852283477783, + "learning_rate": 3.6902065404749006e-05, + "loss": 0.0407, + "step": 36310 + }, + { + "epoch": 1.121056855361442, + "grad_norm": 0.4722125828266144, + "learning_rate": 3.6875471909804516e-05, + "loss": 0.0377, + "step": 36320 + }, + { + "epoch": 1.1213655163899006, + "grad_norm": 0.3899564743041992, + "learning_rate": 3.6848882401745135e-05, + "loss": 0.0391, + "step": 36330 + }, + { + "epoch": 1.1216741774183592, + "grad_norm": 0.46357834339141846, + "learning_rate": 3.682229688864806e-05, + "loss": 0.037, + "step": 36340 + }, + { + "epoch": 1.1219828384468178, + "grad_norm": 0.5799550414085388, + "learning_rate": 3.6795715378589235e-05, + "loss": 0.0422, + "step": 36350 + }, + { + "epoch": 1.1222914994752762, + "grad_norm": 0.3847755193710327, + "learning_rate": 3.676913787964345e-05, + "loss": 0.0336, + "step": 36360 + }, + { + "epoch": 1.1226001605037348, + "grad_norm": 0.2803584933280945, + "learning_rate": 3.674256439988423e-05, + "loss": 0.0437, + "step": 36370 + }, + { + "epoch": 1.1229088215321934, + "grad_norm": 0.29636725783348083, + "learning_rate": 3.6715994947383904e-05, + "loss": 0.0407, + "step": 36380 + }, + { + "epoch": 1.1232174825606518, + "grad_norm": 0.33467572927474976, + "learning_rate": 3.668942953021357e-05, + "loss": 0.0454, + "step": 36390 + }, + { + "epoch": 1.1235261435891104, + "grad_norm": 0.2998700439929962, + "learning_rate": 3.66628681564431e-05, + "loss": 0.0444, + "step": 36400 + }, + { + "epoch": 1.123834804617569, + "grad_norm": 0.29753273725509644, + "learning_rate": 3.663631083414114e-05, + "loss": 0.0396, + "step": 36410 + }, + { + "epoch": 1.1241434656460276, + "grad_norm": 0.3282414674758911, + "learning_rate": 3.660975757137509e-05, + "loss": 0.0434, + "step": 36420 + }, + { + "epoch": 1.124452126674486, + "grad_norm": 0.36831405758857727, + "learning_rate": 3.658320837621114e-05, + "loss": 0.0371, + "step": 36430 + }, + { + "epoch": 1.1247607877029446, + "grad_norm": 0.29213666915893555, + "learning_rate": 3.655666325671426e-05, + "loss": 0.0426, + "step": 36440 + }, + { + "epoch": 1.1250694487314032, + "grad_norm": 0.2714757025241852, + "learning_rate": 3.65301222209481e-05, + "loss": 0.0394, + "step": 36450 + }, + { + "epoch": 1.1253781097598616, + "grad_norm": 0.2683136761188507, + "learning_rate": 3.650358527697519e-05, + "loss": 0.0415, + "step": 36460 + }, + { + "epoch": 1.1256867707883202, + "grad_norm": 0.24732421338558197, + "learning_rate": 3.64770524328567e-05, + "loss": 0.0373, + "step": 36470 + }, + { + "epoch": 1.1259954318167789, + "grad_norm": 0.30757856369018555, + "learning_rate": 3.645052369665265e-05, + "loss": 0.043, + "step": 36480 + }, + { + "epoch": 1.1263040928452375, + "grad_norm": 0.28773197531700134, + "learning_rate": 3.6423999076421724e-05, + "loss": 0.0454, + "step": 36490 + }, + { + "epoch": 1.1266127538736959, + "grad_norm": 0.32750511169433594, + "learning_rate": 3.639747858022142e-05, + "loss": 0.0404, + "step": 36500 + }, + { + "epoch": 1.1269214149021545, + "grad_norm": 0.30749937891960144, + "learning_rate": 3.637096221610799e-05, + "loss": 0.0399, + "step": 36510 + }, + { + "epoch": 1.127230075930613, + "grad_norm": 0.2684897184371948, + "learning_rate": 3.634444999213638e-05, + "loss": 0.0385, + "step": 36520 + }, + { + "epoch": 1.1275387369590715, + "grad_norm": 0.5781752467155457, + "learning_rate": 3.6317941916360296e-05, + "loss": 0.0441, + "step": 36530 + }, + { + "epoch": 1.12784739798753, + "grad_norm": 0.2855963110923767, + "learning_rate": 3.629143799683221e-05, + "loss": 0.0391, + "step": 36540 + }, + { + "epoch": 1.1281560590159887, + "grad_norm": 0.2592746913433075, + "learning_rate": 3.626493824160331e-05, + "loss": 0.0406, + "step": 36550 + }, + { + "epoch": 1.1284647200444473, + "grad_norm": 0.3423193097114563, + "learning_rate": 3.623844265872352e-05, + "loss": 0.0409, + "step": 36560 + }, + { + "epoch": 1.1287733810729057, + "grad_norm": 0.452362596988678, + "learning_rate": 3.621195125624149e-05, + "loss": 0.0394, + "step": 36570 + }, + { + "epoch": 1.1290820421013643, + "grad_norm": 0.47964343428611755, + "learning_rate": 3.618546404220463e-05, + "loss": 0.0402, + "step": 36580 + }, + { + "epoch": 1.1293907031298227, + "grad_norm": 0.34071430563926697, + "learning_rate": 3.615898102465903e-05, + "loss": 0.0444, + "step": 36590 + }, + { + "epoch": 1.1296993641582813, + "grad_norm": 0.3410249650478363, + "learning_rate": 3.6132502211649544e-05, + "loss": 0.0377, + "step": 36600 + }, + { + "epoch": 1.13000802518674, + "grad_norm": 0.3063652217388153, + "learning_rate": 3.610602761121975e-05, + "loss": 0.0382, + "step": 36610 + }, + { + "epoch": 1.1303166862151985, + "grad_norm": 0.44496336579322815, + "learning_rate": 3.6079557231411897e-05, + "loss": 0.0431, + "step": 36620 + }, + { + "epoch": 1.130625347243657, + "grad_norm": 0.377936989068985, + "learning_rate": 3.6053091080267035e-05, + "loss": 0.0424, + "step": 36630 + }, + { + "epoch": 1.1309340082721155, + "grad_norm": 0.3467705547809601, + "learning_rate": 3.602662916582483e-05, + "loss": 0.045, + "step": 36640 + }, + { + "epoch": 1.1312426693005742, + "grad_norm": 0.28116968274116516, + "learning_rate": 3.600017149612375e-05, + "loss": 0.0401, + "step": 36650 + }, + { + "epoch": 1.1315513303290325, + "grad_norm": 0.28235915303230286, + "learning_rate": 3.5973718079200935e-05, + "loss": 0.039, + "step": 36660 + }, + { + "epoch": 1.1318599913574912, + "grad_norm": 0.3686158359050751, + "learning_rate": 3.5947268923092216e-05, + "loss": 0.0383, + "step": 36670 + }, + { + "epoch": 1.1321686523859498, + "grad_norm": 0.28564947843551636, + "learning_rate": 3.592082403583216e-05, + "loss": 0.0374, + "step": 36680 + }, + { + "epoch": 1.1324773134144084, + "grad_norm": 0.3348931074142456, + "learning_rate": 3.5894383425454004e-05, + "loss": 0.0375, + "step": 36690 + }, + { + "epoch": 1.1327859744428668, + "grad_norm": 0.36454805731773376, + "learning_rate": 3.586794709998975e-05, + "loss": 0.0389, + "step": 36700 + }, + { + "epoch": 1.1330946354713254, + "grad_norm": 0.31755173206329346, + "learning_rate": 3.584151506747002e-05, + "loss": 0.0416, + "step": 36710 + }, + { + "epoch": 1.133403296499784, + "grad_norm": 0.2764323651790619, + "learning_rate": 3.581508733592418e-05, + "loss": 0.034, + "step": 36720 + }, + { + "epoch": 1.1337119575282424, + "grad_norm": 0.43975287675857544, + "learning_rate": 3.5788663913380297e-05, + "loss": 0.0431, + "step": 36730 + }, + { + "epoch": 1.134020618556701, + "grad_norm": 0.33344295620918274, + "learning_rate": 3.576224480786506e-05, + "loss": 0.0386, + "step": 36740 + }, + { + "epoch": 1.1343292795851596, + "grad_norm": 0.4283266067504883, + "learning_rate": 3.573583002740393e-05, + "loss": 0.0376, + "step": 36750 + }, + { + "epoch": 1.1346379406136182, + "grad_norm": 0.372330904006958, + "learning_rate": 3.570941958002103e-05, + "loss": 0.0355, + "step": 36760 + }, + { + "epoch": 1.1349466016420766, + "grad_norm": 0.3644862174987793, + "learning_rate": 3.568301347373912e-05, + "loss": 0.0404, + "step": 36770 + }, + { + "epoch": 1.1352552626705352, + "grad_norm": 0.4105653166770935, + "learning_rate": 3.5656611716579726e-05, + "loss": 0.0348, + "step": 36780 + }, + { + "epoch": 1.1355639236989938, + "grad_norm": 0.2960502505302429, + "learning_rate": 3.5630214316562946e-05, + "loss": 0.0403, + "step": 36790 + }, + { + "epoch": 1.1358725847274522, + "grad_norm": 0.3118327260017395, + "learning_rate": 3.560382128170766e-05, + "loss": 0.033, + "step": 36800 + }, + { + "epoch": 1.1361812457559108, + "grad_norm": 0.4107135832309723, + "learning_rate": 3.5577432620031374e-05, + "loss": 0.0431, + "step": 36810 + }, + { + "epoch": 1.1364899067843695, + "grad_norm": 1.0464669466018677, + "learning_rate": 3.5551048339550216e-05, + "loss": 0.0377, + "step": 36820 + }, + { + "epoch": 1.136798567812828, + "grad_norm": 0.3877590596675873, + "learning_rate": 3.55246684482791e-05, + "loss": 0.038, + "step": 36830 + }, + { + "epoch": 1.1371072288412865, + "grad_norm": 0.3598358631134033, + "learning_rate": 3.5498292954231496e-05, + "loss": 0.0367, + "step": 36840 + }, + { + "epoch": 1.137415889869745, + "grad_norm": 0.3011740744113922, + "learning_rate": 3.54719218654196e-05, + "loss": 0.0394, + "step": 36850 + }, + { + "epoch": 1.1377245508982037, + "grad_norm": 0.3727678656578064, + "learning_rate": 3.544555518985425e-05, + "loss": 0.0426, + "step": 36860 + }, + { + "epoch": 1.138033211926662, + "grad_norm": 0.48275038599967957, + "learning_rate": 3.541919293554494e-05, + "loss": 0.042, + "step": 36870 + }, + { + "epoch": 1.1383418729551207, + "grad_norm": 0.3062891662120819, + "learning_rate": 3.539283511049985e-05, + "loss": 0.0386, + "step": 36880 + }, + { + "epoch": 1.1386505339835793, + "grad_norm": 0.285699725151062, + "learning_rate": 3.5366481722725755e-05, + "loss": 0.0397, + "step": 36890 + }, + { + "epoch": 1.1389591950120377, + "grad_norm": 0.4578188955783844, + "learning_rate": 3.534013278022816e-05, + "loss": 0.0389, + "step": 36900 + }, + { + "epoch": 1.1392678560404963, + "grad_norm": 0.34203603863716125, + "learning_rate": 3.531378829101113e-05, + "loss": 0.0377, + "step": 36910 + }, + { + "epoch": 1.139576517068955, + "grad_norm": 0.34845638275146484, + "learning_rate": 3.528744826307746e-05, + "loss": 0.0389, + "step": 36920 + }, + { + "epoch": 1.1398851780974133, + "grad_norm": 0.2808448374271393, + "learning_rate": 3.5261112704428554e-05, + "loss": 0.0376, + "step": 36930 + }, + { + "epoch": 1.140193839125872, + "grad_norm": 0.28144192695617676, + "learning_rate": 3.523478162306443e-05, + "loss": 0.0354, + "step": 36940 + }, + { + "epoch": 1.1405025001543305, + "grad_norm": 0.36442166566848755, + "learning_rate": 3.520845502698381e-05, + "loss": 0.0393, + "step": 36950 + }, + { + "epoch": 1.1408111611827891, + "grad_norm": 0.30765873193740845, + "learning_rate": 3.5182132924184005e-05, + "loss": 0.0366, + "step": 36960 + }, + { + "epoch": 1.1411198222112475, + "grad_norm": 0.3164180815219879, + "learning_rate": 3.5155815322660966e-05, + "loss": 0.0402, + "step": 36970 + }, + { + "epoch": 1.1414284832397061, + "grad_norm": 0.3145584464073181, + "learning_rate": 3.512950223040931e-05, + "loss": 0.0398, + "step": 36980 + }, + { + "epoch": 1.1417371442681647, + "grad_norm": 0.307148277759552, + "learning_rate": 3.5103193655422216e-05, + "loss": 0.0355, + "step": 36990 + }, + { + "epoch": 1.1420458052966231, + "grad_norm": 0.3261582851409912, + "learning_rate": 3.5076889605691596e-05, + "loss": 0.0389, + "step": 37000 + }, + { + "epoch": 1.1423544663250818, + "grad_norm": 0.34509822726249695, + "learning_rate": 3.505059008920787e-05, + "loss": 0.0414, + "step": 37010 + }, + { + "epoch": 1.1426631273535404, + "grad_norm": 0.26995849609375, + "learning_rate": 3.502429511396016e-05, + "loss": 0.0355, + "step": 37020 + }, + { + "epoch": 1.142971788381999, + "grad_norm": 0.27754032611846924, + "learning_rate": 3.4998004687936196e-05, + "loss": 0.039, + "step": 37030 + }, + { + "epoch": 1.1432804494104574, + "grad_norm": 0.21456609666347504, + "learning_rate": 3.497171881912229e-05, + "loss": 0.0448, + "step": 37040 + }, + { + "epoch": 1.143589110438916, + "grad_norm": 0.303864061832428, + "learning_rate": 3.494543751550342e-05, + "loss": 0.0403, + "step": 37050 + }, + { + "epoch": 1.1438977714673746, + "grad_norm": 0.5106635689735413, + "learning_rate": 3.491916078506313e-05, + "loss": 0.0415, + "step": 37060 + }, + { + "epoch": 1.144206432495833, + "grad_norm": 0.3778795003890991, + "learning_rate": 3.489288863578361e-05, + "loss": 0.0449, + "step": 37070 + }, + { + "epoch": 1.1445150935242916, + "grad_norm": 0.5484660267829895, + "learning_rate": 3.4866621075645646e-05, + "loss": 0.0389, + "step": 37080 + }, + { + "epoch": 1.1448237545527502, + "grad_norm": 0.31577208638191223, + "learning_rate": 3.4840358112628614e-05, + "loss": 0.04, + "step": 37090 + }, + { + "epoch": 1.1451324155812088, + "grad_norm": 0.24930912256240845, + "learning_rate": 3.481409975471053e-05, + "loss": 0.0414, + "step": 37100 + }, + { + "epoch": 1.1454410766096672, + "grad_norm": 0.3237205743789673, + "learning_rate": 3.4787846009867986e-05, + "loss": 0.036, + "step": 37110 + }, + { + "epoch": 1.1457497376381258, + "grad_norm": 0.3394230902194977, + "learning_rate": 3.476159688607615e-05, + "loss": 0.0416, + "step": 37120 + }, + { + "epoch": 1.1460583986665844, + "grad_norm": 0.28779226541519165, + "learning_rate": 3.4735352391308854e-05, + "loss": 0.0404, + "step": 37130 + }, + { + "epoch": 1.1463670596950428, + "grad_norm": 0.3849194645881653, + "learning_rate": 3.4709112533538446e-05, + "loss": 0.038, + "step": 37140 + }, + { + "epoch": 1.1466757207235014, + "grad_norm": 0.5817273259162903, + "learning_rate": 3.4682877320735934e-05, + "loss": 0.0404, + "step": 37150 + }, + { + "epoch": 1.14698438175196, + "grad_norm": 0.3499465882778168, + "learning_rate": 3.465664676087085e-05, + "loss": 0.0418, + "step": 37160 + }, + { + "epoch": 1.1472930427804187, + "grad_norm": 0.28933992981910706, + "learning_rate": 3.463042086191136e-05, + "loss": 0.0376, + "step": 37170 + }, + { + "epoch": 1.147601703808877, + "grad_norm": 0.5414829254150391, + "learning_rate": 3.460419963182423e-05, + "loss": 0.0403, + "step": 37180 + }, + { + "epoch": 1.1479103648373357, + "grad_norm": 0.38319462537765503, + "learning_rate": 3.457798307857473e-05, + "loss": 0.0428, + "step": 37190 + }, + { + "epoch": 1.1482190258657943, + "grad_norm": 0.32183924317359924, + "learning_rate": 3.455177121012678e-05, + "loss": 0.0371, + "step": 37200 + }, + { + "epoch": 1.1485276868942527, + "grad_norm": 0.38008883595466614, + "learning_rate": 3.452556403444285e-05, + "loss": 0.0387, + "step": 37210 + }, + { + "epoch": 1.1488363479227113, + "grad_norm": 0.3078245520591736, + "learning_rate": 3.4499361559483975e-05, + "loss": 0.0394, + "step": 37220 + }, + { + "epoch": 1.1491450089511699, + "grad_norm": 0.49804458022117615, + "learning_rate": 3.44731637932098e-05, + "loss": 0.0393, + "step": 37230 + }, + { + "epoch": 1.1494536699796283, + "grad_norm": 0.3046203851699829, + "learning_rate": 3.44469707435785e-05, + "loss": 0.0396, + "step": 37240 + }, + { + "epoch": 1.149762331008087, + "grad_norm": 0.3634600043296814, + "learning_rate": 3.4420782418546835e-05, + "loss": 0.0373, + "step": 37250 + }, + { + "epoch": 1.1500709920365455, + "grad_norm": 0.41060417890548706, + "learning_rate": 3.439459882607012e-05, + "loss": 0.0429, + "step": 37260 + }, + { + "epoch": 1.150379653065004, + "grad_norm": 0.27803975343704224, + "learning_rate": 3.436841997410225e-05, + "loss": 0.04, + "step": 37270 + }, + { + "epoch": 1.1506883140934625, + "grad_norm": 0.3598780333995819, + "learning_rate": 3.434224587059567e-05, + "loss": 0.0394, + "step": 37280 + }, + { + "epoch": 1.1509969751219211, + "grad_norm": 0.4561542868614197, + "learning_rate": 3.431607652350136e-05, + "loss": 0.0431, + "step": 37290 + }, + { + "epoch": 1.1513056361503797, + "grad_norm": 0.4273194372653961, + "learning_rate": 3.428991194076891e-05, + "loss": 0.0382, + "step": 37300 + }, + { + "epoch": 1.1516142971788381, + "grad_norm": 0.3289939761161804, + "learning_rate": 3.4263752130346394e-05, + "loss": 0.039, + "step": 37310 + }, + { + "epoch": 1.1519229582072967, + "grad_norm": 0.43834301829338074, + "learning_rate": 3.4237597100180515e-05, + "loss": 0.0394, + "step": 37320 + }, + { + "epoch": 1.1522316192357553, + "grad_norm": 0.6011613011360168, + "learning_rate": 3.4211446858216427e-05, + "loss": 0.0397, + "step": 37330 + }, + { + "epoch": 1.1525402802642137, + "grad_norm": 0.3358350694179535, + "learning_rate": 3.4185301412397915e-05, + "loss": 0.0417, + "step": 37340 + }, + { + "epoch": 1.1528489412926723, + "grad_norm": 0.4039250314235687, + "learning_rate": 3.415916077066729e-05, + "loss": 0.0404, + "step": 37350 + }, + { + "epoch": 1.153157602321131, + "grad_norm": 0.360161155462265, + "learning_rate": 3.413302494096535e-05, + "loss": 0.0379, + "step": 37360 + }, + { + "epoch": 1.1534662633495896, + "grad_norm": 0.34873726963996887, + "learning_rate": 3.410689393123151e-05, + "loss": 0.0382, + "step": 37370 + }, + { + "epoch": 1.153774924378048, + "grad_norm": 0.3667021095752716, + "learning_rate": 3.408076774940364e-05, + "loss": 0.0401, + "step": 37380 + }, + { + "epoch": 1.1540835854065066, + "grad_norm": 0.41862472891807556, + "learning_rate": 3.40546464034182e-05, + "loss": 0.0364, + "step": 37390 + }, + { + "epoch": 1.1543922464349652, + "grad_norm": 0.2977229356765747, + "learning_rate": 3.4028529901210185e-05, + "loss": 0.0395, + "step": 37400 + }, + { + "epoch": 1.1547009074634236, + "grad_norm": 0.3083047568798065, + "learning_rate": 3.4002418250713086e-05, + "loss": 0.0407, + "step": 37410 + }, + { + "epoch": 1.1550095684918822, + "grad_norm": 0.41021281480789185, + "learning_rate": 3.3976311459858936e-05, + "loss": 0.0437, + "step": 37420 + }, + { + "epoch": 1.1553182295203408, + "grad_norm": 0.5623751282691956, + "learning_rate": 3.395020953657826e-05, + "loss": 0.037, + "step": 37430 + }, + { + "epoch": 1.1556268905487994, + "grad_norm": 0.26422709226608276, + "learning_rate": 3.3924112488800165e-05, + "loss": 0.0393, + "step": 37440 + }, + { + "epoch": 1.1559355515772578, + "grad_norm": 0.2759140431880951, + "learning_rate": 3.389802032445225e-05, + "loss": 0.0423, + "step": 37450 + }, + { + "epoch": 1.1562442126057164, + "grad_norm": 0.252689927816391, + "learning_rate": 3.38719330514606e-05, + "loss": 0.0408, + "step": 37460 + }, + { + "epoch": 1.156552873634175, + "grad_norm": 0.3662818670272827, + "learning_rate": 3.3845850677749866e-05, + "loss": 0.0382, + "step": 37470 + }, + { + "epoch": 1.1568615346626334, + "grad_norm": 0.40958458185195923, + "learning_rate": 3.3819773211243157e-05, + "loss": 0.0353, + "step": 37480 + }, + { + "epoch": 1.157170195691092, + "grad_norm": 0.3181270360946655, + "learning_rate": 3.379370065986213e-05, + "loss": 0.0408, + "step": 37490 + }, + { + "epoch": 1.1574788567195506, + "grad_norm": 0.3758452832698822, + "learning_rate": 3.3767633031526955e-05, + "loss": 0.0436, + "step": 37500 + }, + { + "epoch": 1.1577875177480093, + "grad_norm": 0.3234855532646179, + "learning_rate": 3.374157033415626e-05, + "loss": 0.0455, + "step": 37510 + }, + { + "epoch": 1.1580961787764676, + "grad_norm": 0.28279751539230347, + "learning_rate": 3.371551257566723e-05, + "loss": 0.0418, + "step": 37520 + }, + { + "epoch": 1.1584048398049263, + "grad_norm": 0.41119977831840515, + "learning_rate": 3.36894597639755e-05, + "loss": 0.0394, + "step": 37530 + }, + { + "epoch": 1.1587135008333849, + "grad_norm": 0.29867491126060486, + "learning_rate": 3.366341190699523e-05, + "loss": 0.0427, + "step": 37540 + }, + { + "epoch": 1.1590221618618433, + "grad_norm": 0.34491801261901855, + "learning_rate": 3.36373690126391e-05, + "loss": 0.0388, + "step": 37550 + }, + { + "epoch": 1.1593308228903019, + "grad_norm": 0.4054747223854065, + "learning_rate": 3.3611331088818234e-05, + "loss": 0.0419, + "step": 37560 + }, + { + "epoch": 1.1596394839187605, + "grad_norm": 0.38076063990592957, + "learning_rate": 3.3585298143442265e-05, + "loss": 0.0424, + "step": 37570 + }, + { + "epoch": 1.1599481449472189, + "grad_norm": 0.32387444376945496, + "learning_rate": 3.35592701844193e-05, + "loss": 0.0375, + "step": 37580 + }, + { + "epoch": 1.1602568059756775, + "grad_norm": 0.26610037684440613, + "learning_rate": 3.353324721965596e-05, + "loss": 0.0407, + "step": 37590 + }, + { + "epoch": 1.160565467004136, + "grad_norm": 0.2949923574924469, + "learning_rate": 3.350722925705736e-05, + "loss": 0.0452, + "step": 37600 + }, + { + "epoch": 1.1608741280325945, + "grad_norm": 0.35733968019485474, + "learning_rate": 3.348121630452703e-05, + "loss": 0.0375, + "step": 37610 + }, + { + "epoch": 1.161182789061053, + "grad_norm": 0.4990885555744171, + "learning_rate": 3.3455208369967044e-05, + "loss": 0.0404, + "step": 37620 + }, + { + "epoch": 1.1614914500895117, + "grad_norm": 0.32246139645576477, + "learning_rate": 3.34292054612779e-05, + "loss": 0.0389, + "step": 37630 + }, + { + "epoch": 1.1618001111179703, + "grad_norm": 0.3078879117965698, + "learning_rate": 3.340320758635861e-05, + "loss": 0.0396, + "step": 37640 + }, + { + "epoch": 1.1621087721464287, + "grad_norm": 0.365713894367218, + "learning_rate": 3.337721475310666e-05, + "loss": 0.0418, + "step": 37650 + }, + { + "epoch": 1.1624174331748873, + "grad_norm": 0.3639325499534607, + "learning_rate": 3.335122696941795e-05, + "loss": 0.0394, + "step": 37660 + }, + { + "epoch": 1.162726094203346, + "grad_norm": 0.2601913511753082, + "learning_rate": 3.332524424318692e-05, + "loss": 0.0358, + "step": 37670 + }, + { + "epoch": 1.1630347552318043, + "grad_norm": 0.4648866057395935, + "learning_rate": 3.32992665823064e-05, + "loss": 0.0369, + "step": 37680 + }, + { + "epoch": 1.163343416260263, + "grad_norm": 0.4743565022945404, + "learning_rate": 3.327329399466774e-05, + "loss": 0.0398, + "step": 37690 + }, + { + "epoch": 1.1636520772887216, + "grad_norm": 0.425370454788208, + "learning_rate": 3.324732648816072e-05, + "loss": 0.0393, + "step": 37700 + }, + { + "epoch": 1.1639607383171802, + "grad_norm": 0.3596694767475128, + "learning_rate": 3.322136407067358e-05, + "loss": 0.0425, + "step": 37710 + }, + { + "epoch": 1.1642693993456386, + "grad_norm": 0.3161143958568573, + "learning_rate": 3.3195406750093036e-05, + "loss": 0.0361, + "step": 37720 + }, + { + "epoch": 1.1645780603740972, + "grad_norm": 0.40972253680229187, + "learning_rate": 3.3169454534304205e-05, + "loss": 0.0385, + "step": 37730 + }, + { + "epoch": 1.1648867214025558, + "grad_norm": 0.2819901406764984, + "learning_rate": 3.3143507431190725e-05, + "loss": 0.0495, + "step": 37740 + }, + { + "epoch": 1.1651953824310142, + "grad_norm": 0.33006903529167175, + "learning_rate": 3.311756544863459e-05, + "loss": 0.0426, + "step": 37750 + }, + { + "epoch": 1.1655040434594728, + "grad_norm": 0.4026828706264496, + "learning_rate": 3.309162859451633e-05, + "loss": 0.0392, + "step": 37760 + }, + { + "epoch": 1.1658127044879314, + "grad_norm": 0.3356577754020691, + "learning_rate": 3.306569687671487e-05, + "loss": 0.0395, + "step": 37770 + }, + { + "epoch": 1.16612136551639, + "grad_norm": 0.3912312984466553, + "learning_rate": 3.303977030310756e-05, + "loss": 0.0394, + "step": 37780 + }, + { + "epoch": 1.1664300265448484, + "grad_norm": 0.27019834518432617, + "learning_rate": 3.3013848881570245e-05, + "loss": 0.0384, + "step": 37790 + }, + { + "epoch": 1.166738687573307, + "grad_norm": 0.37219807505607605, + "learning_rate": 3.298793261997712e-05, + "loss": 0.0386, + "step": 37800 + }, + { + "epoch": 1.1670473486017656, + "grad_norm": 0.27471837401390076, + "learning_rate": 3.2962021526200893e-05, + "loss": 0.0384, + "step": 37810 + }, + { + "epoch": 1.167356009630224, + "grad_norm": 0.2922915816307068, + "learning_rate": 3.293611560811268e-05, + "loss": 0.0395, + "step": 37820 + }, + { + "epoch": 1.1676646706586826, + "grad_norm": 0.3804088830947876, + "learning_rate": 3.291021487358199e-05, + "loss": 0.0399, + "step": 37830 + }, + { + "epoch": 1.1679733316871412, + "grad_norm": 0.30952632427215576, + "learning_rate": 3.28843193304768e-05, + "loss": 0.0381, + "step": 37840 + }, + { + "epoch": 1.1682819927155998, + "grad_norm": 0.2785918116569519, + "learning_rate": 3.2858428986663456e-05, + "loss": 0.0377, + "step": 37850 + }, + { + "epoch": 1.1685906537440582, + "grad_norm": 0.3624306917190552, + "learning_rate": 3.283254385000681e-05, + "loss": 0.0401, + "step": 37860 + }, + { + "epoch": 1.1688993147725169, + "grad_norm": 0.40940913558006287, + "learning_rate": 3.2806663928370076e-05, + "loss": 0.0407, + "step": 37870 + }, + { + "epoch": 1.1692079758009755, + "grad_norm": 0.3131282925605774, + "learning_rate": 3.278078922961485e-05, + "loss": 0.0422, + "step": 37880 + }, + { + "epoch": 1.1695166368294339, + "grad_norm": 0.2917909324169159, + "learning_rate": 3.275491976160123e-05, + "loss": 0.0352, + "step": 37890 + }, + { + "epoch": 1.1698252978578925, + "grad_norm": 0.38550248742103577, + "learning_rate": 3.2729055532187645e-05, + "loss": 0.0395, + "step": 37900 + }, + { + "epoch": 1.170133958886351, + "grad_norm": 0.29250332713127136, + "learning_rate": 3.270319654923097e-05, + "loss": 0.037, + "step": 37910 + }, + { + "epoch": 1.1704426199148095, + "grad_norm": 0.30632826685905457, + "learning_rate": 3.2677342820586506e-05, + "loss": 0.0425, + "step": 37920 + }, + { + "epoch": 1.170751280943268, + "grad_norm": 0.37874311208724976, + "learning_rate": 3.2651494354107905e-05, + "loss": 0.0439, + "step": 37930 + }, + { + "epoch": 1.1710599419717267, + "grad_norm": 0.4418604075908661, + "learning_rate": 3.2625651157647266e-05, + "loss": 0.0364, + "step": 37940 + }, + { + "epoch": 1.171368603000185, + "grad_norm": 0.43150264024734497, + "learning_rate": 3.259981323905505e-05, + "loss": 0.0379, + "step": 37950 + }, + { + "epoch": 1.1716772640286437, + "grad_norm": 0.332766056060791, + "learning_rate": 3.257398060618014e-05, + "loss": 0.0398, + "step": 37960 + }, + { + "epoch": 1.1719859250571023, + "grad_norm": 0.45333874225616455, + "learning_rate": 3.254815326686983e-05, + "loss": 0.0406, + "step": 37970 + }, + { + "epoch": 1.172294586085561, + "grad_norm": 0.3994063138961792, + "learning_rate": 3.2522331228969774e-05, + "loss": 0.0376, + "step": 37980 + }, + { + "epoch": 1.1726032471140193, + "grad_norm": 0.3103809654712677, + "learning_rate": 3.2496514500324006e-05, + "loss": 0.0376, + "step": 37990 + }, + { + "epoch": 1.172911908142478, + "grad_norm": 0.2964399456977844, + "learning_rate": 3.247070308877498e-05, + "loss": 0.0355, + "step": 38000 + }, + { + "epoch": 1.1732205691709365, + "grad_norm": 0.3811151087284088, + "learning_rate": 3.2444897002163515e-05, + "loss": 0.0338, + "step": 38010 + }, + { + "epoch": 1.173529230199395, + "grad_norm": 0.4024549424648285, + "learning_rate": 3.241909624832885e-05, + "loss": 0.0386, + "step": 38020 + }, + { + "epoch": 1.1738378912278535, + "grad_norm": 0.2615504562854767, + "learning_rate": 3.239330083510852e-05, + "loss": 0.037, + "step": 38030 + }, + { + "epoch": 1.1741465522563121, + "grad_norm": 0.3622654378414154, + "learning_rate": 3.236751077033855e-05, + "loss": 0.0369, + "step": 38040 + }, + { + "epoch": 1.1744552132847708, + "grad_norm": 0.23131319880485535, + "learning_rate": 3.234172606185322e-05, + "loss": 0.0385, + "step": 38050 + }, + { + "epoch": 1.1747638743132292, + "grad_norm": 0.29487067461013794, + "learning_rate": 3.231594671748528e-05, + "loss": 0.0428, + "step": 38060 + }, + { + "epoch": 1.1750725353416878, + "grad_norm": 0.5425028800964355, + "learning_rate": 3.2290172745065815e-05, + "loss": 0.0418, + "step": 38070 + }, + { + "epoch": 1.1753811963701464, + "grad_norm": 0.35272863507270813, + "learning_rate": 3.226440415242426e-05, + "loss": 0.0428, + "step": 38080 + }, + { + "epoch": 1.1756898573986048, + "grad_norm": 0.3553423285484314, + "learning_rate": 3.223864094738846e-05, + "loss": 0.041, + "step": 38090 + }, + { + "epoch": 1.1759985184270634, + "grad_norm": 0.3768143951892853, + "learning_rate": 3.221288313778456e-05, + "loss": 0.0351, + "step": 38100 + }, + { + "epoch": 1.176307179455522, + "grad_norm": 0.3607240319252014, + "learning_rate": 3.2187130731437125e-05, + "loss": 0.0387, + "step": 38110 + }, + { + "epoch": 1.1766158404839806, + "grad_norm": 0.29928916692733765, + "learning_rate": 3.216138373616905e-05, + "loss": 0.0383, + "step": 38120 + }, + { + "epoch": 1.176924501512439, + "grad_norm": 0.30244702100753784, + "learning_rate": 3.21356421598016e-05, + "loss": 0.0385, + "step": 38130 + }, + { + "epoch": 1.1772331625408976, + "grad_norm": 0.3601962625980377, + "learning_rate": 3.210990601015438e-05, + "loss": 0.036, + "step": 38140 + }, + { + "epoch": 1.1775418235693562, + "grad_norm": 0.5427117347717285, + "learning_rate": 3.208417529504535e-05, + "loss": 0.0458, + "step": 38150 + }, + { + "epoch": 1.1778504845978146, + "grad_norm": 0.3282106816768646, + "learning_rate": 3.205845002229084e-05, + "loss": 0.0383, + "step": 38160 + }, + { + "epoch": 1.1781591456262732, + "grad_norm": 0.348327100276947, + "learning_rate": 3.203273019970547e-05, + "loss": 0.0393, + "step": 38170 + }, + { + "epoch": 1.1784678066547318, + "grad_norm": 0.34963226318359375, + "learning_rate": 3.200701583510227e-05, + "loss": 0.0392, + "step": 38180 + }, + { + "epoch": 1.1787764676831904, + "grad_norm": 0.5467879772186279, + "learning_rate": 3.198130693629261e-05, + "loss": 0.04, + "step": 38190 + }, + { + "epoch": 1.1790851287116488, + "grad_norm": 0.2740291357040405, + "learning_rate": 3.195560351108612e-05, + "loss": 0.0352, + "step": 38200 + }, + { + "epoch": 1.1793937897401074, + "grad_norm": 0.4135095477104187, + "learning_rate": 3.1929905567290865e-05, + "loss": 0.0417, + "step": 38210 + }, + { + "epoch": 1.179702450768566, + "grad_norm": 0.3895938992500305, + "learning_rate": 3.1904213112713164e-05, + "loss": 0.0357, + "step": 38220 + }, + { + "epoch": 1.1800111117970244, + "grad_norm": 0.41168391704559326, + "learning_rate": 3.187852615515774e-05, + "loss": 0.0368, + "step": 38230 + }, + { + "epoch": 1.180319772825483, + "grad_norm": 0.283019095659256, + "learning_rate": 3.1852844702427606e-05, + "loss": 0.0355, + "step": 38240 + }, + { + "epoch": 1.1806284338539417, + "grad_norm": 0.2794211506843567, + "learning_rate": 3.18271687623241e-05, + "loss": 0.0347, + "step": 38250 + }, + { + "epoch": 1.1809370948824, + "grad_norm": 0.33059605956077576, + "learning_rate": 3.1801498342646896e-05, + "loss": 0.0389, + "step": 38260 + }, + { + "epoch": 1.1812457559108587, + "grad_norm": 0.35933172702789307, + "learning_rate": 3.177583345119398e-05, + "loss": 0.0398, + "step": 38270 + }, + { + "epoch": 1.1815544169393173, + "grad_norm": 0.31388962268829346, + "learning_rate": 3.17501740957617e-05, + "loss": 0.0384, + "step": 38280 + }, + { + "epoch": 1.1818630779677757, + "grad_norm": 0.232137069106102, + "learning_rate": 3.172452028414467e-05, + "loss": 0.0401, + "step": 38290 + }, + { + "epoch": 1.1821717389962343, + "grad_norm": 0.34333115816116333, + "learning_rate": 3.169887202413583e-05, + "loss": 0.0425, + "step": 38300 + }, + { + "epoch": 1.182480400024693, + "grad_norm": 0.36816659569740295, + "learning_rate": 3.167322932352646e-05, + "loss": 0.0378, + "step": 38310 + }, + { + "epoch": 1.1827890610531515, + "grad_norm": 0.3450593650341034, + "learning_rate": 3.164759219010613e-05, + "loss": 0.0382, + "step": 38320 + }, + { + "epoch": 1.18309772208161, + "grad_norm": 0.33669790625572205, + "learning_rate": 3.1621960631662725e-05, + "loss": 0.0375, + "step": 38330 + }, + { + "epoch": 1.1834063831100685, + "grad_norm": 0.243030846118927, + "learning_rate": 3.159633465598245e-05, + "loss": 0.0421, + "step": 38340 + }, + { + "epoch": 1.1837150441385271, + "grad_norm": 0.3108178377151489, + "learning_rate": 3.1570714270849767e-05, + "loss": 0.0381, + "step": 38350 + }, + { + "epoch": 1.1840237051669855, + "grad_norm": 0.43112048506736755, + "learning_rate": 3.1545099484047516e-05, + "loss": 0.0349, + "step": 38360 + }, + { + "epoch": 1.1843323661954441, + "grad_norm": 0.3354644775390625, + "learning_rate": 3.151949030335674e-05, + "loss": 0.0356, + "step": 38370 + }, + { + "epoch": 1.1846410272239027, + "grad_norm": 0.39079129695892334, + "learning_rate": 3.149388673655687e-05, + "loss": 0.0341, + "step": 38380 + }, + { + "epoch": 1.1849496882523614, + "grad_norm": 0.266806423664093, + "learning_rate": 3.146828879142559e-05, + "loss": 0.042, + "step": 38390 + }, + { + "epoch": 1.1852583492808197, + "grad_norm": 0.4852360785007477, + "learning_rate": 3.1442696475738866e-05, + "loss": 0.0388, + "step": 38400 + }, + { + "epoch": 1.1855670103092784, + "grad_norm": 0.5405202507972717, + "learning_rate": 3.141710979727098e-05, + "loss": 0.0344, + "step": 38410 + }, + { + "epoch": 1.185875671337737, + "grad_norm": 0.3281932771205902, + "learning_rate": 3.139152876379447e-05, + "loss": 0.0374, + "step": 38420 + }, + { + "epoch": 1.1861843323661954, + "grad_norm": 0.37637728452682495, + "learning_rate": 3.1365953383080214e-05, + "loss": 0.0367, + "step": 38430 + }, + { + "epoch": 1.186492993394654, + "grad_norm": 0.33621746301651, + "learning_rate": 3.134038366289731e-05, + "loss": 0.038, + "step": 38440 + }, + { + "epoch": 1.1868016544231126, + "grad_norm": 0.3079611659049988, + "learning_rate": 3.131481961101317e-05, + "loss": 0.0439, + "step": 38450 + }, + { + "epoch": 1.1871103154515712, + "grad_norm": 0.28574660420417786, + "learning_rate": 3.128926123519349e-05, + "loss": 0.0378, + "step": 38460 + }, + { + "epoch": 1.1874189764800296, + "grad_norm": 0.511269211769104, + "learning_rate": 3.1263708543202194e-05, + "loss": 0.0392, + "step": 38470 + }, + { + "epoch": 1.1877276375084882, + "grad_norm": 0.30802321434020996, + "learning_rate": 3.123816154280155e-05, + "loss": 0.0388, + "step": 38480 + }, + { + "epoch": 1.1880362985369468, + "grad_norm": 0.35515594482421875, + "learning_rate": 3.121262024175207e-05, + "loss": 0.0377, + "step": 38490 + }, + { + "epoch": 1.1883449595654052, + "grad_norm": 0.34856075048446655, + "learning_rate": 3.118708464781248e-05, + "loss": 0.036, + "step": 38500 + }, + { + "epoch": 1.1886536205938638, + "grad_norm": 0.5380772948265076, + "learning_rate": 3.116155476873987e-05, + "loss": 0.0361, + "step": 38510 + }, + { + "epoch": 1.1889622816223224, + "grad_norm": 0.3129236698150635, + "learning_rate": 3.11360306122895e-05, + "loss": 0.0425, + "step": 38520 + }, + { + "epoch": 1.1892709426507808, + "grad_norm": 0.3877452611923218, + "learning_rate": 3.1110512186214975e-05, + "loss": 0.0408, + "step": 38530 + }, + { + "epoch": 1.1895796036792394, + "grad_norm": 0.8704223036766052, + "learning_rate": 3.1084999498268095e-05, + "loss": 0.0385, + "step": 38540 + }, + { + "epoch": 1.189888264707698, + "grad_norm": 0.32816460728645325, + "learning_rate": 3.1059492556198934e-05, + "loss": 0.0399, + "step": 38550 + }, + { + "epoch": 1.1901969257361564, + "grad_norm": 0.3146388530731201, + "learning_rate": 3.103399136775586e-05, + "loss": 0.0393, + "step": 38560 + }, + { + "epoch": 1.190505586764615, + "grad_norm": 0.40355512499809265, + "learning_rate": 3.100849594068541e-05, + "loss": 0.0418, + "step": 38570 + }, + { + "epoch": 1.1908142477930737, + "grad_norm": 0.26714590191841125, + "learning_rate": 3.0983006282732484e-05, + "loss": 0.036, + "step": 38580 + }, + { + "epoch": 1.1911229088215323, + "grad_norm": 0.3359220623970032, + "learning_rate": 3.0957522401640116e-05, + "loss": 0.0421, + "step": 38590 + }, + { + "epoch": 1.1914315698499907, + "grad_norm": 0.4440728724002838, + "learning_rate": 3.0932044305149645e-05, + "loss": 0.0402, + "step": 38600 + }, + { + "epoch": 1.1917402308784493, + "grad_norm": 0.2611095905303955, + "learning_rate": 3.090657200100068e-05, + "loss": 0.042, + "step": 38610 + }, + { + "epoch": 1.1920488919069079, + "grad_norm": 0.8410347700119019, + "learning_rate": 3.088110549693099e-05, + "loss": 0.0385, + "step": 38620 + }, + { + "epoch": 1.1923575529353663, + "grad_norm": 0.2505684792995453, + "learning_rate": 3.085564480067667e-05, + "loss": 0.0358, + "step": 38630 + }, + { + "epoch": 1.1926662139638249, + "grad_norm": 0.39697355031967163, + "learning_rate": 3.0830189919971955e-05, + "loss": 0.0385, + "step": 38640 + }, + { + "epoch": 1.1929748749922835, + "grad_norm": 0.3052166998386383, + "learning_rate": 3.080474086254939e-05, + "loss": 0.0422, + "step": 38650 + }, + { + "epoch": 1.193283536020742, + "grad_norm": 0.40931737422943115, + "learning_rate": 3.077929763613975e-05, + "loss": 0.039, + "step": 38660 + }, + { + "epoch": 1.1935921970492005, + "grad_norm": 0.26873940229415894, + "learning_rate": 3.075386024847198e-05, + "loss": 0.0374, + "step": 38670 + }, + { + "epoch": 1.193900858077659, + "grad_norm": 0.3048810660839081, + "learning_rate": 3.072842870727331e-05, + "loss": 0.0381, + "step": 38680 + }, + { + "epoch": 1.1942095191061177, + "grad_norm": 0.3726906180381775, + "learning_rate": 3.070300302026916e-05, + "loss": 0.0344, + "step": 38690 + }, + { + "epoch": 1.1945181801345761, + "grad_norm": 0.39980852603912354, + "learning_rate": 3.067758319518318e-05, + "loss": 0.0408, + "step": 38700 + }, + { + "epoch": 1.1948268411630347, + "grad_norm": 0.40129995346069336, + "learning_rate": 3.065216923973725e-05, + "loss": 0.0371, + "step": 38710 + }, + { + "epoch": 1.1951355021914933, + "grad_norm": 0.28942129015922546, + "learning_rate": 3.062676116165145e-05, + "loss": 0.0383, + "step": 38720 + }, + { + "epoch": 1.195444163219952, + "grad_norm": 0.246199369430542, + "learning_rate": 3.06013589686441e-05, + "loss": 0.0413, + "step": 38730 + }, + { + "epoch": 1.1957528242484103, + "grad_norm": 0.25981616973876953, + "learning_rate": 3.05759626684317e-05, + "loss": 0.0418, + "step": 38740 + }, + { + "epoch": 1.196061485276869, + "grad_norm": 0.30037549138069153, + "learning_rate": 3.055057226872896e-05, + "loss": 0.04, + "step": 38750 + }, + { + "epoch": 1.1963701463053276, + "grad_norm": 0.26539722084999084, + "learning_rate": 3.052518777724887e-05, + "loss": 0.0392, + "step": 38760 + }, + { + "epoch": 1.196678807333786, + "grad_norm": 0.3359214961528778, + "learning_rate": 3.04998092017025e-05, + "loss": 0.0371, + "step": 38770 + }, + { + "epoch": 1.1969874683622446, + "grad_norm": 0.3542978763580322, + "learning_rate": 3.0474436549799246e-05, + "loss": 0.0388, + "step": 38780 + }, + { + "epoch": 1.1972961293907032, + "grad_norm": 0.24360202252864838, + "learning_rate": 3.044906982924661e-05, + "loss": 0.0378, + "step": 38790 + }, + { + "epoch": 1.1976047904191618, + "grad_norm": 0.28930938243865967, + "learning_rate": 3.0423709047750337e-05, + "loss": 0.036, + "step": 38800 + }, + { + "epoch": 1.1979134514476202, + "grad_norm": 0.36919867992401123, + "learning_rate": 3.03983542130144e-05, + "loss": 0.0392, + "step": 38810 + }, + { + "epoch": 1.1982221124760788, + "grad_norm": 0.28380995988845825, + "learning_rate": 3.0373005332740877e-05, + "loss": 0.0374, + "step": 38820 + }, + { + "epoch": 1.1985307735045374, + "grad_norm": 0.3404821455478668, + "learning_rate": 3.034766241463013e-05, + "loss": 0.0398, + "step": 38830 + }, + { + "epoch": 1.1988394345329958, + "grad_norm": 0.3478001654148102, + "learning_rate": 3.032232546638064e-05, + "loss": 0.037, + "step": 38840 + }, + { + "epoch": 1.1991480955614544, + "grad_norm": 0.32412686944007874, + "learning_rate": 3.0296994495689114e-05, + "loss": 0.0374, + "step": 38850 + }, + { + "epoch": 1.199456756589913, + "grad_norm": 0.3145819306373596, + "learning_rate": 3.0271669510250444e-05, + "loss": 0.0352, + "step": 38860 + }, + { + "epoch": 1.1997654176183714, + "grad_norm": 0.32861652970314026, + "learning_rate": 3.024635051775766e-05, + "loss": 0.0388, + "step": 38870 + }, + { + "epoch": 1.20007407864683, + "grad_norm": 0.4090190827846527, + "learning_rate": 3.022103752590205e-05, + "loss": 0.0411, + "step": 38880 + }, + { + "epoch": 1.2003827396752886, + "grad_norm": 0.24426577985286713, + "learning_rate": 3.0195730542372992e-05, + "loss": 0.0371, + "step": 38890 + }, + { + "epoch": 1.200691400703747, + "grad_norm": 0.3027993440628052, + "learning_rate": 3.0170429574858084e-05, + "loss": 0.0388, + "step": 38900 + }, + { + "epoch": 1.2010000617322056, + "grad_norm": 0.5191834568977356, + "learning_rate": 3.0145134631043127e-05, + "loss": 0.0412, + "step": 38910 + }, + { + "epoch": 1.2013087227606642, + "grad_norm": 0.4062039256095886, + "learning_rate": 3.0119845718612018e-05, + "loss": 0.0376, + "step": 38920 + }, + { + "epoch": 1.2016173837891229, + "grad_norm": 0.43583840131759644, + "learning_rate": 3.009456284524688e-05, + "loss": 0.0411, + "step": 38930 + }, + { + "epoch": 1.2019260448175813, + "grad_norm": 0.3071868121623993, + "learning_rate": 3.0069286018627967e-05, + "loss": 0.0357, + "step": 38940 + }, + { + "epoch": 1.2022347058460399, + "grad_norm": 0.4569222331047058, + "learning_rate": 3.0044015246433743e-05, + "loss": 0.0361, + "step": 38950 + }, + { + "epoch": 1.2025433668744985, + "grad_norm": 0.3426021933555603, + "learning_rate": 3.0018750536340755e-05, + "loss": 0.0358, + "step": 38960 + }, + { + "epoch": 1.2028520279029569, + "grad_norm": 0.40307649970054626, + "learning_rate": 2.999349189602378e-05, + "loss": 0.0358, + "step": 38970 + }, + { + "epoch": 1.2031606889314155, + "grad_norm": 0.30242758989334106, + "learning_rate": 2.9968239333155733e-05, + "loss": 0.0343, + "step": 38980 + }, + { + "epoch": 1.203469349959874, + "grad_norm": 0.3507860600948334, + "learning_rate": 2.994299285540767e-05, + "loss": 0.0435, + "step": 38990 + }, + { + "epoch": 1.2037780109883327, + "grad_norm": 0.24192693829536438, + "learning_rate": 2.9917752470448813e-05, + "loss": 0.0373, + "step": 39000 + }, + { + "epoch": 1.204086672016791, + "grad_norm": 0.2993853688240051, + "learning_rate": 2.9892518185946495e-05, + "loss": 0.0381, + "step": 39010 + }, + { + "epoch": 1.2043953330452497, + "grad_norm": 0.38233065605163574, + "learning_rate": 2.986729000956624e-05, + "loss": 0.0368, + "step": 39020 + }, + { + "epoch": 1.2047039940737083, + "grad_norm": 0.26018425822257996, + "learning_rate": 2.9842067948971736e-05, + "loss": 0.0385, + "step": 39030 + }, + { + "epoch": 1.2050126551021667, + "grad_norm": 0.43633079528808594, + "learning_rate": 2.9816852011824727e-05, + "loss": 0.0406, + "step": 39040 + }, + { + "epoch": 1.2053213161306253, + "grad_norm": 0.2606615424156189, + "learning_rate": 2.979164220578519e-05, + "loss": 0.0368, + "step": 39050 + }, + { + "epoch": 1.205629977159084, + "grad_norm": 0.2774304747581482, + "learning_rate": 2.9766438538511165e-05, + "loss": 0.039, + "step": 39060 + }, + { + "epoch": 1.2059386381875425, + "grad_norm": 0.33742669224739075, + "learning_rate": 2.9741241017658873e-05, + "loss": 0.0366, + "step": 39070 + }, + { + "epoch": 1.206247299216001, + "grad_norm": 0.6431738138198853, + "learning_rate": 2.971604965088267e-05, + "loss": 0.0387, + "step": 39080 + }, + { + "epoch": 1.2065559602444595, + "grad_norm": 0.3000437021255493, + "learning_rate": 2.9690864445835008e-05, + "loss": 0.0393, + "step": 39090 + }, + { + "epoch": 1.2068646212729182, + "grad_norm": 0.4202556014060974, + "learning_rate": 2.966568541016651e-05, + "loss": 0.0372, + "step": 39100 + }, + { + "epoch": 1.2071732823013765, + "grad_norm": 0.4385071396827698, + "learning_rate": 2.9640512551525867e-05, + "loss": 0.0377, + "step": 39110 + }, + { + "epoch": 1.2074819433298352, + "grad_norm": 0.32261890172958374, + "learning_rate": 2.961534587755995e-05, + "loss": 0.0391, + "step": 39120 + }, + { + "epoch": 1.2077906043582938, + "grad_norm": 0.296415239572525, + "learning_rate": 2.959018539591375e-05, + "loss": 0.0381, + "step": 39130 + }, + { + "epoch": 1.2080992653867524, + "grad_norm": 0.3654562532901764, + "learning_rate": 2.9565031114230325e-05, + "loss": 0.037, + "step": 39140 + }, + { + "epoch": 1.2084079264152108, + "grad_norm": 0.47445201873779297, + "learning_rate": 2.9539883040150895e-05, + "loss": 0.037, + "step": 39150 + }, + { + "epoch": 1.2087165874436694, + "grad_norm": 0.33166465163230896, + "learning_rate": 2.9514741181314774e-05, + "loss": 0.0394, + "step": 39160 + }, + { + "epoch": 1.209025248472128, + "grad_norm": 0.28900742530822754, + "learning_rate": 2.94896055453594e-05, + "loss": 0.0383, + "step": 39170 + }, + { + "epoch": 1.2093339095005864, + "grad_norm": 0.36310872435569763, + "learning_rate": 2.9464476139920332e-05, + "loss": 0.0382, + "step": 39180 + }, + { + "epoch": 1.209642570529045, + "grad_norm": 0.43638941645622253, + "learning_rate": 2.9439352972631186e-05, + "loss": 0.0394, + "step": 39190 + }, + { + "epoch": 1.2099512315575036, + "grad_norm": 0.31224656105041504, + "learning_rate": 2.9414236051123757e-05, + "loss": 0.0382, + "step": 39200 + }, + { + "epoch": 1.210259892585962, + "grad_norm": 0.38435012102127075, + "learning_rate": 2.938912538302785e-05, + "loss": 0.0369, + "step": 39210 + }, + { + "epoch": 1.2105685536144206, + "grad_norm": 0.31209996342658997, + "learning_rate": 2.9364020975971464e-05, + "loss": 0.0394, + "step": 39220 + }, + { + "epoch": 1.2108772146428792, + "grad_norm": 0.28802451491355896, + "learning_rate": 2.9338922837580657e-05, + "loss": 0.0393, + "step": 39230 + }, + { + "epoch": 1.2111858756713376, + "grad_norm": 0.3961065709590912, + "learning_rate": 2.931383097547955e-05, + "loss": 0.0393, + "step": 39240 + }, + { + "epoch": 1.2114945366997962, + "grad_norm": 0.3731761574745178, + "learning_rate": 2.928874539729043e-05, + "loss": 0.0383, + "step": 39250 + }, + { + "epoch": 1.2118031977282548, + "grad_norm": 0.2798191010951996, + "learning_rate": 2.926366611063358e-05, + "loss": 0.0412, + "step": 39260 + }, + { + "epoch": 1.2121118587567135, + "grad_norm": 0.379838764667511, + "learning_rate": 2.9238593123127463e-05, + "loss": 0.0399, + "step": 39270 + }, + { + "epoch": 1.2124205197851718, + "grad_norm": 0.37955453991889954, + "learning_rate": 2.9213526442388583e-05, + "loss": 0.0369, + "step": 39280 + }, + { + "epoch": 1.2127291808136305, + "grad_norm": 0.3506087064743042, + "learning_rate": 2.9188466076031545e-05, + "loss": 0.036, + "step": 39290 + }, + { + "epoch": 1.213037841842089, + "grad_norm": 0.31766897439956665, + "learning_rate": 2.9163412031669012e-05, + "loss": 0.0317, + "step": 39300 + }, + { + "epoch": 1.2133465028705475, + "grad_norm": 0.2936810851097107, + "learning_rate": 2.913836431691175e-05, + "loss": 0.0453, + "step": 39310 + }, + { + "epoch": 1.213655163899006, + "grad_norm": 0.25011345744132996, + "learning_rate": 2.9113322939368583e-05, + "loss": 0.0376, + "step": 39320 + }, + { + "epoch": 1.2139638249274647, + "grad_norm": 0.3263258635997772, + "learning_rate": 2.9088287906646427e-05, + "loss": 0.0395, + "step": 39330 + }, + { + "epoch": 1.2142724859559233, + "grad_norm": 0.3032049834728241, + "learning_rate": 2.906325922635024e-05, + "loss": 0.0435, + "step": 39340 + }, + { + "epoch": 1.2145811469843817, + "grad_norm": 0.36757954955101013, + "learning_rate": 2.903823690608313e-05, + "loss": 0.0391, + "step": 39350 + }, + { + "epoch": 1.2148898080128403, + "grad_norm": 0.33083346486091614, + "learning_rate": 2.9013220953446174e-05, + "loss": 0.0393, + "step": 39360 + }, + { + "epoch": 1.215198469041299, + "grad_norm": 0.2903749644756317, + "learning_rate": 2.8988211376038564e-05, + "loss": 0.0425, + "step": 39370 + }, + { + "epoch": 1.2155071300697573, + "grad_norm": 0.2424953728914261, + "learning_rate": 2.8963208181457564e-05, + "loss": 0.0398, + "step": 39380 + }, + { + "epoch": 1.215815791098216, + "grad_norm": 0.2849297523498535, + "learning_rate": 2.8938211377298453e-05, + "loss": 0.0387, + "step": 39390 + }, + { + "epoch": 1.2161244521266745, + "grad_norm": 0.37069669365882874, + "learning_rate": 2.8913220971154652e-05, + "loss": 0.0403, + "step": 39400 + }, + { + "epoch": 1.2164331131551331, + "grad_norm": 0.3311381936073303, + "learning_rate": 2.888823697061753e-05, + "loss": 0.0386, + "step": 39410 + }, + { + "epoch": 1.2167417741835915, + "grad_norm": 0.3765101134777069, + "learning_rate": 2.8863259383276618e-05, + "loss": 0.0432, + "step": 39420 + }, + { + "epoch": 1.2170504352120501, + "grad_norm": 0.40129992365837097, + "learning_rate": 2.8838288216719395e-05, + "loss": 0.037, + "step": 39430 + }, + { + "epoch": 1.2173590962405088, + "grad_norm": 0.34476566314697266, + "learning_rate": 2.8813323478531484e-05, + "loss": 0.037, + "step": 39440 + }, + { + "epoch": 1.2176677572689671, + "grad_norm": 0.40650177001953125, + "learning_rate": 2.8788365176296496e-05, + "loss": 0.0418, + "step": 39450 + }, + { + "epoch": 1.2179764182974258, + "grad_norm": 0.44825926423072815, + "learning_rate": 2.876341331759611e-05, + "loss": 0.0399, + "step": 39460 + }, + { + "epoch": 1.2182850793258844, + "grad_norm": 0.32021844387054443, + "learning_rate": 2.8738467910010036e-05, + "loss": 0.035, + "step": 39470 + }, + { + "epoch": 1.218593740354343, + "grad_norm": 0.35859426856040955, + "learning_rate": 2.8713528961116032e-05, + "loss": 0.0385, + "step": 39480 + }, + { + "epoch": 1.2189024013828014, + "grad_norm": 0.5193864107131958, + "learning_rate": 2.8688596478489875e-05, + "loss": 0.0354, + "step": 39490 + }, + { + "epoch": 1.21921106241126, + "grad_norm": 0.28135645389556885, + "learning_rate": 2.8663670469705434e-05, + "loss": 0.0357, + "step": 39500 + }, + { + "epoch": 1.2195197234397186, + "grad_norm": 0.3918088972568512, + "learning_rate": 2.8638750942334546e-05, + "loss": 0.0406, + "step": 39510 + }, + { + "epoch": 1.219828384468177, + "grad_norm": 0.4055735468864441, + "learning_rate": 2.8613837903947115e-05, + "loss": 0.0335, + "step": 39520 + }, + { + "epoch": 1.2201370454966356, + "grad_norm": 0.3266981840133667, + "learning_rate": 2.858893136211106e-05, + "loss": 0.0375, + "step": 39530 + }, + { + "epoch": 1.2204457065250942, + "grad_norm": 0.3116229772567749, + "learning_rate": 2.8564031324392315e-05, + "loss": 0.042, + "step": 39540 + }, + { + "epoch": 1.2207543675535526, + "grad_norm": 0.3429498076438904, + "learning_rate": 2.85391377983549e-05, + "loss": 0.0356, + "step": 39550 + }, + { + "epoch": 1.2210630285820112, + "grad_norm": 0.3500301241874695, + "learning_rate": 2.851425079156075e-05, + "loss": 0.0385, + "step": 39560 + }, + { + "epoch": 1.2213716896104698, + "grad_norm": 0.40213876962661743, + "learning_rate": 2.848937031156994e-05, + "loss": 0.0451, + "step": 39570 + }, + { + "epoch": 1.2216803506389282, + "grad_norm": 0.21777793765068054, + "learning_rate": 2.846449636594044e-05, + "loss": 0.0332, + "step": 39580 + }, + { + "epoch": 1.2219890116673868, + "grad_norm": 0.2889273464679718, + "learning_rate": 2.843962896222836e-05, + "loss": 0.0331, + "step": 39590 + }, + { + "epoch": 1.2222976726958454, + "grad_norm": 0.3597586154937744, + "learning_rate": 2.8414768107987722e-05, + "loss": 0.0392, + "step": 39600 + }, + { + "epoch": 1.222606333724304, + "grad_norm": 0.3697020709514618, + "learning_rate": 2.838991381077061e-05, + "loss": 0.0396, + "step": 39610 + }, + { + "epoch": 1.2229149947527624, + "grad_norm": 0.3256124258041382, + "learning_rate": 2.83650660781271e-05, + "loss": 0.0397, + "step": 39620 + }, + { + "epoch": 1.223223655781221, + "grad_norm": 0.27182894945144653, + "learning_rate": 2.8340224917605285e-05, + "loss": 0.0354, + "step": 39630 + }, + { + "epoch": 1.2235323168096797, + "grad_norm": 0.3003523051738739, + "learning_rate": 2.831539033675122e-05, + "loss": 0.0387, + "step": 39640 + }, + { + "epoch": 1.223840977838138, + "grad_norm": 0.3977314233779907, + "learning_rate": 2.8290562343109038e-05, + "loss": 0.0396, + "step": 39650 + }, + { + "epoch": 1.2241496388665967, + "grad_norm": 0.4731379449367523, + "learning_rate": 2.826574094422082e-05, + "loss": 0.0382, + "step": 39660 + }, + { + "epoch": 1.2244582998950553, + "grad_norm": 0.5647000074386597, + "learning_rate": 2.8240926147626645e-05, + "loss": 0.0337, + "step": 39670 + }, + { + "epoch": 1.224766960923514, + "grad_norm": 0.32192471623420715, + "learning_rate": 2.8216117960864586e-05, + "loss": 0.0373, + "step": 39680 + }, + { + "epoch": 1.2250756219519723, + "grad_norm": 0.3306925594806671, + "learning_rate": 2.8191316391470703e-05, + "loss": 0.0427, + "step": 39690 + }, + { + "epoch": 1.225384282980431, + "grad_norm": 0.3308136463165283, + "learning_rate": 2.816652144697911e-05, + "loss": 0.0381, + "step": 39700 + }, + { + "epoch": 1.2256929440088895, + "grad_norm": 0.34869351983070374, + "learning_rate": 2.8141733134921783e-05, + "loss": 0.0393, + "step": 39710 + }, + { + "epoch": 1.226001605037348, + "grad_norm": 0.3059689402580261, + "learning_rate": 2.811695146282884e-05, + "loss": 0.0377, + "step": 39720 + }, + { + "epoch": 1.2263102660658065, + "grad_norm": 0.36837053298950195, + "learning_rate": 2.8092176438228212e-05, + "loss": 0.0394, + "step": 39730 + }, + { + "epoch": 1.2266189270942651, + "grad_norm": 0.3908098042011261, + "learning_rate": 2.806740806864598e-05, + "loss": 0.0387, + "step": 39740 + }, + { + "epoch": 1.2269275881227237, + "grad_norm": 0.43140217661857605, + "learning_rate": 2.804264636160604e-05, + "loss": 0.0398, + "step": 39750 + }, + { + "epoch": 1.2272362491511821, + "grad_norm": 0.33969566226005554, + "learning_rate": 2.8017891324630402e-05, + "loss": 0.0379, + "step": 39760 + }, + { + "epoch": 1.2275449101796407, + "grad_norm": 0.49882057309150696, + "learning_rate": 2.7993142965238976e-05, + "loss": 0.0394, + "step": 39770 + }, + { + "epoch": 1.2278535712080993, + "grad_norm": 0.3256571888923645, + "learning_rate": 2.7968401290949665e-05, + "loss": 0.0366, + "step": 39780 + }, + { + "epoch": 1.2281622322365577, + "grad_norm": 0.34735947847366333, + "learning_rate": 2.7943666309278328e-05, + "loss": 0.0397, + "step": 39790 + }, + { + "epoch": 1.2284708932650164, + "grad_norm": 0.41716423630714417, + "learning_rate": 2.7918938027738783e-05, + "loss": 0.0392, + "step": 39800 + }, + { + "epoch": 1.228779554293475, + "grad_norm": 0.36087772250175476, + "learning_rate": 2.789421645384287e-05, + "loss": 0.0357, + "step": 39810 + }, + { + "epoch": 1.2290882153219336, + "grad_norm": 0.4724888503551483, + "learning_rate": 2.786950159510032e-05, + "loss": 0.0401, + "step": 39820 + }, + { + "epoch": 1.229396876350392, + "grad_norm": 0.28516319394111633, + "learning_rate": 2.7844793459018876e-05, + "loss": 0.0366, + "step": 39830 + }, + { + "epoch": 1.2297055373788506, + "grad_norm": 0.34144166111946106, + "learning_rate": 2.7820092053104195e-05, + "loss": 0.0371, + "step": 39840 + }, + { + "epoch": 1.2300141984073092, + "grad_norm": 0.3376360535621643, + "learning_rate": 2.7795397384859933e-05, + "loss": 0.0422, + "step": 39850 + }, + { + "epoch": 1.2303228594357676, + "grad_norm": 0.23520833253860474, + "learning_rate": 2.7770709461787638e-05, + "loss": 0.0368, + "step": 39860 + }, + { + "epoch": 1.2306315204642262, + "grad_norm": 0.3424001634120941, + "learning_rate": 2.7746028291386915e-05, + "loss": 0.0367, + "step": 39870 + }, + { + "epoch": 1.2309401814926848, + "grad_norm": 0.3947562575340271, + "learning_rate": 2.772135388115519e-05, + "loss": 0.0415, + "step": 39880 + }, + { + "epoch": 1.2312488425211432, + "grad_norm": 0.3141418695449829, + "learning_rate": 2.7696686238587945e-05, + "loss": 0.0361, + "step": 39890 + }, + { + "epoch": 1.2315575035496018, + "grad_norm": 0.34930163621902466, + "learning_rate": 2.7672025371178505e-05, + "loss": 0.0391, + "step": 39900 + }, + { + "epoch": 1.2318661645780604, + "grad_norm": 0.31554755568504333, + "learning_rate": 2.7647371286418238e-05, + "loss": 0.0368, + "step": 39910 + }, + { + "epoch": 1.2321748256065188, + "grad_norm": 0.28790682554244995, + "learning_rate": 2.762272399179639e-05, + "loss": 0.037, + "step": 39920 + }, + { + "epoch": 1.2324834866349774, + "grad_norm": 0.3265746235847473, + "learning_rate": 2.7598083494800154e-05, + "loss": 0.0388, + "step": 39930 + }, + { + "epoch": 1.232792147663436, + "grad_norm": 0.34355348348617554, + "learning_rate": 2.7573449802914664e-05, + "loss": 0.0327, + "step": 39940 + }, + { + "epoch": 1.2331008086918946, + "grad_norm": 0.5224756598472595, + "learning_rate": 2.7548822923622964e-05, + "loss": 0.0361, + "step": 39950 + }, + { + "epoch": 1.233409469720353, + "grad_norm": 0.29086101055145264, + "learning_rate": 2.752420286440609e-05, + "loss": 0.0387, + "step": 39960 + }, + { + "epoch": 1.2337181307488116, + "grad_norm": 0.3530241549015045, + "learning_rate": 2.749958963274295e-05, + "loss": 0.0375, + "step": 39970 + }, + { + "epoch": 1.2340267917772703, + "grad_norm": 0.3077525198459625, + "learning_rate": 2.747498323611039e-05, + "loss": 0.0389, + "step": 39980 + }, + { + "epoch": 1.2343354528057287, + "grad_norm": 0.38159406185150146, + "learning_rate": 2.7450383681983184e-05, + "loss": 0.0398, + "step": 39990 + }, + { + "epoch": 1.2346441138341873, + "grad_norm": 0.21740418672561646, + "learning_rate": 2.742579097783403e-05, + "loss": 0.0407, + "step": 40000 + }, + { + "epoch": 1.2349527748626459, + "grad_norm": 0.5631418228149414, + "learning_rate": 2.7401205131133512e-05, + "loss": 0.0392, + "step": 40010 + }, + { + "epoch": 1.2352614358911045, + "grad_norm": 0.29256635904312134, + "learning_rate": 2.7376626149350238e-05, + "loss": 0.0368, + "step": 40020 + }, + { + "epoch": 1.2355700969195629, + "grad_norm": 0.4207102656364441, + "learning_rate": 2.735205403995056e-05, + "loss": 0.0409, + "step": 40030 + }, + { + "epoch": 1.2358787579480215, + "grad_norm": 0.28802141547203064, + "learning_rate": 2.7327488810398917e-05, + "loss": 0.0402, + "step": 40040 + }, + { + "epoch": 1.23618741897648, + "grad_norm": 0.7137340903282166, + "learning_rate": 2.7302930468157507e-05, + "loss": 0.0383, + "step": 40050 + }, + { + "epoch": 1.2364960800049385, + "grad_norm": 0.6111975908279419, + "learning_rate": 2.727837902068655e-05, + "loss": 0.0373, + "step": 40060 + }, + { + "epoch": 1.236804741033397, + "grad_norm": 0.27809837460517883, + "learning_rate": 2.7253834475444123e-05, + "loss": 0.0374, + "step": 40070 + }, + { + "epoch": 1.2371134020618557, + "grad_norm": 0.27012479305267334, + "learning_rate": 2.7229296839886204e-05, + "loss": 0.0439, + "step": 40080 + }, + { + "epoch": 1.2374220630903143, + "grad_norm": 0.30687305331230164, + "learning_rate": 2.720476612146668e-05, + "loss": 0.039, + "step": 40090 + }, + { + "epoch": 1.2377307241187727, + "grad_norm": 0.26958751678466797, + "learning_rate": 2.7180242327637317e-05, + "loss": 0.0394, + "step": 40100 + }, + { + "epoch": 1.2380393851472313, + "grad_norm": 0.3943692147731781, + "learning_rate": 2.7155725465847826e-05, + "loss": 0.0362, + "step": 40110 + }, + { + "epoch": 1.23834804617569, + "grad_norm": 0.3672271966934204, + "learning_rate": 2.713121554354578e-05, + "loss": 0.0345, + "step": 40120 + }, + { + "epoch": 1.2386567072041483, + "grad_norm": 0.30341315269470215, + "learning_rate": 2.7106712568176628e-05, + "loss": 0.0394, + "step": 40130 + }, + { + "epoch": 1.238965368232607, + "grad_norm": 0.27626127004623413, + "learning_rate": 2.708221654718374e-05, + "loss": 0.0396, + "step": 40140 + }, + { + "epoch": 1.2392740292610656, + "grad_norm": 0.3237275779247284, + "learning_rate": 2.7057727488008357e-05, + "loss": 0.0421, + "step": 40150 + }, + { + "epoch": 1.2395826902895242, + "grad_norm": 0.2572384476661682, + "learning_rate": 2.703324539808961e-05, + "loss": 0.0384, + "step": 40160 + }, + { + "epoch": 1.2398913513179826, + "grad_norm": 0.22608023881912231, + "learning_rate": 2.7008770284864505e-05, + "loss": 0.0416, + "step": 40170 + }, + { + "epoch": 1.2402000123464412, + "grad_norm": 0.4029091000556946, + "learning_rate": 2.6984302155767916e-05, + "loss": 0.0334, + "step": 40180 + }, + { + "epoch": 1.2405086733748996, + "grad_norm": 0.41810086369514465, + "learning_rate": 2.6959841018232683e-05, + "loss": 0.0392, + "step": 40190 + }, + { + "epoch": 1.2408173344033582, + "grad_norm": 0.4083022177219391, + "learning_rate": 2.693538687968937e-05, + "loss": 0.035, + "step": 40200 + }, + { + "epoch": 1.2411259954318168, + "grad_norm": 0.3098403513431549, + "learning_rate": 2.6910939747566556e-05, + "loss": 0.0373, + "step": 40210 + }, + { + "epoch": 1.2414346564602754, + "grad_norm": 0.4225579500198364, + "learning_rate": 2.6886499629290607e-05, + "loss": 0.0381, + "step": 40220 + }, + { + "epoch": 1.2417433174887338, + "grad_norm": 0.3457123339176178, + "learning_rate": 2.6862066532285802e-05, + "loss": 0.0341, + "step": 40230 + }, + { + "epoch": 1.2420519785171924, + "grad_norm": 0.40356677770614624, + "learning_rate": 2.6837640463974262e-05, + "loss": 0.0355, + "step": 40240 + }, + { + "epoch": 1.242360639545651, + "grad_norm": 0.34220054745674133, + "learning_rate": 2.681322143177596e-05, + "loss": 0.0419, + "step": 40250 + }, + { + "epoch": 1.2426693005741094, + "grad_norm": 0.31409788131713867, + "learning_rate": 2.678880944310882e-05, + "loss": 0.0359, + "step": 40260 + }, + { + "epoch": 1.242977961602568, + "grad_norm": 0.29770371317863464, + "learning_rate": 2.6764404505388474e-05, + "loss": 0.038, + "step": 40270 + }, + { + "epoch": 1.2432866226310266, + "grad_norm": 0.47271931171417236, + "learning_rate": 2.6740006626028558e-05, + "loss": 0.0362, + "step": 40280 + }, + { + "epoch": 1.2435952836594852, + "grad_norm": 0.48012715578079224, + "learning_rate": 2.671561581244048e-05, + "loss": 0.0385, + "step": 40290 + }, + { + "epoch": 1.2439039446879436, + "grad_norm": 0.3193557560443878, + "learning_rate": 2.6691232072033536e-05, + "loss": 0.0434, + "step": 40300 + }, + { + "epoch": 1.2442126057164022, + "grad_norm": 0.24300603568553925, + "learning_rate": 2.6666855412214852e-05, + "loss": 0.0415, + "step": 40310 + }, + { + "epoch": 1.2445212667448609, + "grad_norm": 0.5098138451576233, + "learning_rate": 2.664248584038942e-05, + "loss": 0.0325, + "step": 40320 + }, + { + "epoch": 1.2448299277733192, + "grad_norm": 0.2893952429294586, + "learning_rate": 2.6618123363960047e-05, + "loss": 0.0382, + "step": 40330 + }, + { + "epoch": 1.2451385888017779, + "grad_norm": 0.5158940553665161, + "learning_rate": 2.659376799032748e-05, + "loss": 0.036, + "step": 40340 + }, + { + "epoch": 1.2454472498302365, + "grad_norm": 0.31645023822784424, + "learning_rate": 2.6569419726890145e-05, + "loss": 0.042, + "step": 40350 + }, + { + "epoch": 1.245755910858695, + "grad_norm": 0.4164888560771942, + "learning_rate": 2.654507858104447e-05, + "loss": 0.0392, + "step": 40360 + }, + { + "epoch": 1.2460645718871535, + "grad_norm": 0.32525813579559326, + "learning_rate": 2.652074456018463e-05, + "loss": 0.0396, + "step": 40370 + }, + { + "epoch": 1.246373232915612, + "grad_norm": 0.533382773399353, + "learning_rate": 2.6496417671702646e-05, + "loss": 0.0367, + "step": 40380 + }, + { + "epoch": 1.2466818939440707, + "grad_norm": 0.4478584825992584, + "learning_rate": 2.6472097922988427e-05, + "loss": 0.0372, + "step": 40390 + }, + { + "epoch": 1.246990554972529, + "grad_norm": 0.4205123484134674, + "learning_rate": 2.6447785321429607e-05, + "loss": 0.0369, + "step": 40400 + }, + { + "epoch": 1.2472992160009877, + "grad_norm": 0.321258008480072, + "learning_rate": 2.6423479874411784e-05, + "loss": 0.0393, + "step": 40410 + }, + { + "epoch": 1.2476078770294463, + "grad_norm": 0.3933238685131073, + "learning_rate": 2.6399181589318234e-05, + "loss": 0.0349, + "step": 40420 + }, + { + "epoch": 1.247916538057905, + "grad_norm": 0.4252876043319702, + "learning_rate": 2.6374890473530188e-05, + "loss": 0.0417, + "step": 40430 + }, + { + "epoch": 1.2482251990863633, + "grad_norm": 0.4361044466495514, + "learning_rate": 2.635060653442664e-05, + "loss": 0.0387, + "step": 40440 + }, + { + "epoch": 1.248533860114822, + "grad_norm": 0.6562917828559875, + "learning_rate": 2.6326329779384395e-05, + "loss": 0.036, + "step": 40450 + }, + { + "epoch": 1.2488425211432805, + "grad_norm": 0.34543824195861816, + "learning_rate": 2.63020602157781e-05, + "loss": 0.0388, + "step": 40460 + }, + { + "epoch": 1.249151182171739, + "grad_norm": 0.26832714676856995, + "learning_rate": 2.62777978509802e-05, + "loss": 0.0386, + "step": 40470 + }, + { + "epoch": 1.2494598432001975, + "grad_norm": 0.3014635741710663, + "learning_rate": 2.6253542692360954e-05, + "loss": 0.0384, + "step": 40480 + }, + { + "epoch": 1.2497685042286562, + "grad_norm": 0.5039165019989014, + "learning_rate": 2.6229294747288458e-05, + "loss": 0.039, + "step": 40490 + }, + { + "epoch": 1.2500771652571148, + "grad_norm": 0.3546570837497711, + "learning_rate": 2.6205054023128596e-05, + "loss": 0.0372, + "step": 40500 + }, + { + "epoch": 1.2503858262855732, + "grad_norm": 0.5112490057945251, + "learning_rate": 2.6180820527245043e-05, + "loss": 0.038, + "step": 40510 + }, + { + "epoch": 1.2506944873140318, + "grad_norm": 0.27835434675216675, + "learning_rate": 2.6156594266999313e-05, + "loss": 0.037, + "step": 40520 + }, + { + "epoch": 1.2510031483424902, + "grad_norm": 0.29021674394607544, + "learning_rate": 2.6132375249750672e-05, + "loss": 0.0373, + "step": 40530 + }, + { + "epoch": 1.2513118093709488, + "grad_norm": 0.2711077630519867, + "learning_rate": 2.6108163482856286e-05, + "loss": 0.0427, + "step": 40540 + }, + { + "epoch": 1.2516204703994074, + "grad_norm": 0.3214209973812103, + "learning_rate": 2.6083958973670964e-05, + "loss": 0.0345, + "step": 40550 + }, + { + "epoch": 1.251929131427866, + "grad_norm": 0.43092843890190125, + "learning_rate": 2.6059761729547483e-05, + "loss": 0.0367, + "step": 40560 + }, + { + "epoch": 1.2522377924563244, + "grad_norm": 0.3391376733779907, + "learning_rate": 2.603557175783624e-05, + "loss": 0.0398, + "step": 40570 + }, + { + "epoch": 1.252546453484783, + "grad_norm": 0.3132045865058899, + "learning_rate": 2.601138906588559e-05, + "loss": 0.0339, + "step": 40580 + }, + { + "epoch": 1.2528551145132416, + "grad_norm": 0.631365954875946, + "learning_rate": 2.598721366104152e-05, + "loss": 0.037, + "step": 40590 + }, + { + "epoch": 1.2531637755417, + "grad_norm": 0.409047931432724, + "learning_rate": 2.5963045550647945e-05, + "loss": 0.0423, + "step": 40600 + }, + { + "epoch": 1.2534724365701586, + "grad_norm": 0.32197296619415283, + "learning_rate": 2.5938884742046466e-05, + "loss": 0.0404, + "step": 40610 + }, + { + "epoch": 1.2537810975986172, + "grad_norm": 0.33429643511772156, + "learning_rate": 2.5914731242576507e-05, + "loss": 0.0427, + "step": 40620 + }, + { + "epoch": 1.2540897586270758, + "grad_norm": 0.44775447249412537, + "learning_rate": 2.5890585059575268e-05, + "loss": 0.0351, + "step": 40630 + }, + { + "epoch": 1.2543984196555342, + "grad_norm": 0.36595913767814636, + "learning_rate": 2.5866446200377688e-05, + "loss": 0.0404, + "step": 40640 + }, + { + "epoch": 1.2547070806839928, + "grad_norm": 0.39842090010643005, + "learning_rate": 2.5842314672316566e-05, + "loss": 0.0341, + "step": 40650 + }, + { + "epoch": 1.2550157417124514, + "grad_norm": 0.5650822520256042, + "learning_rate": 2.581819048272239e-05, + "loss": 0.0375, + "step": 40660 + }, + { + "epoch": 1.2553244027409098, + "grad_norm": 0.31089383363723755, + "learning_rate": 2.5794073638923478e-05, + "loss": 0.0339, + "step": 40670 + }, + { + "epoch": 1.2556330637693685, + "grad_norm": 0.24984395503997803, + "learning_rate": 2.576996414824586e-05, + "loss": 0.0383, + "step": 40680 + }, + { + "epoch": 1.255941724797827, + "grad_norm": 0.27803024649620056, + "learning_rate": 2.574586201801339e-05, + "loss": 0.0393, + "step": 40690 + }, + { + "epoch": 1.2562503858262857, + "grad_norm": 0.3400859236717224, + "learning_rate": 2.572176725554762e-05, + "loss": 0.0387, + "step": 40700 + }, + { + "epoch": 1.256559046854744, + "grad_norm": 0.4205976724624634, + "learning_rate": 2.5697679868167966e-05, + "loss": 0.04, + "step": 40710 + }, + { + "epoch": 1.2568677078832027, + "grad_norm": 0.24525168538093567, + "learning_rate": 2.5673599863191468e-05, + "loss": 0.0377, + "step": 40720 + }, + { + "epoch": 1.2571763689116613, + "grad_norm": 0.3199567198753357, + "learning_rate": 2.564952724793306e-05, + "loss": 0.0366, + "step": 40730 + }, + { + "epoch": 1.2574850299401197, + "grad_norm": 0.25972771644592285, + "learning_rate": 2.5625462029705306e-05, + "loss": 0.0365, + "step": 40740 + }, + { + "epoch": 1.2577936909685783, + "grad_norm": 0.5670916438102722, + "learning_rate": 2.5601404215818624e-05, + "loss": 0.0359, + "step": 40750 + }, + { + "epoch": 1.258102351997037, + "grad_norm": 0.29433107376098633, + "learning_rate": 2.5577353813581144e-05, + "loss": 0.0463, + "step": 40760 + }, + { + "epoch": 1.2584110130254955, + "grad_norm": 0.424870103597641, + "learning_rate": 2.5553310830298733e-05, + "loss": 0.0384, + "step": 40770 + }, + { + "epoch": 1.258719674053954, + "grad_norm": 0.41980403661727905, + "learning_rate": 2.5529275273275012e-05, + "loss": 0.0401, + "step": 40780 + }, + { + "epoch": 1.2590283350824125, + "grad_norm": 0.29325827956199646, + "learning_rate": 2.550524714981133e-05, + "loss": 0.0354, + "step": 40790 + }, + { + "epoch": 1.259336996110871, + "grad_norm": 0.32255974411964417, + "learning_rate": 2.5481226467206837e-05, + "loss": 0.0367, + "step": 40800 + }, + { + "epoch": 1.2596456571393295, + "grad_norm": 0.25812453031539917, + "learning_rate": 2.5457213232758365e-05, + "loss": 0.0371, + "step": 40810 + }, + { + "epoch": 1.2599543181677881, + "grad_norm": 0.3666881322860718, + "learning_rate": 2.5433207453760498e-05, + "loss": 0.0398, + "step": 40820 + }, + { + "epoch": 1.2602629791962467, + "grad_norm": 0.38005855679512024, + "learning_rate": 2.5409209137505552e-05, + "loss": 0.0393, + "step": 40830 + }, + { + "epoch": 1.2605716402247054, + "grad_norm": 0.36154183745384216, + "learning_rate": 2.5385218291283597e-05, + "loss": 0.0368, + "step": 40840 + }, + { + "epoch": 1.2608803012531637, + "grad_norm": 0.25393176078796387, + "learning_rate": 2.5361234922382383e-05, + "loss": 0.0368, + "step": 40850 + }, + { + "epoch": 1.2611889622816224, + "grad_norm": 0.3130289614200592, + "learning_rate": 2.533725903808749e-05, + "loss": 0.0346, + "step": 40860 + }, + { + "epoch": 1.2614976233100808, + "grad_norm": 0.2720595598220825, + "learning_rate": 2.5313290645682085e-05, + "loss": 0.0392, + "step": 40870 + }, + { + "epoch": 1.2618062843385394, + "grad_norm": 0.37836408615112305, + "learning_rate": 2.52893297524472e-05, + "loss": 0.0419, + "step": 40880 + }, + { + "epoch": 1.262114945366998, + "grad_norm": 0.4368501305580139, + "learning_rate": 2.526537636566145e-05, + "loss": 0.0413, + "step": 40890 + }, + { + "epoch": 1.2624236063954566, + "grad_norm": 0.30789706110954285, + "learning_rate": 2.5241430492601305e-05, + "loss": 0.0356, + "step": 40900 + }, + { + "epoch": 1.262732267423915, + "grad_norm": 0.3324950933456421, + "learning_rate": 2.5217492140540867e-05, + "loss": 0.0404, + "step": 40910 + }, + { + "epoch": 1.2630409284523736, + "grad_norm": 0.6906617879867554, + "learning_rate": 2.5193561316751967e-05, + "loss": 0.0355, + "step": 40920 + }, + { + "epoch": 1.2633495894808322, + "grad_norm": 0.26258745789527893, + "learning_rate": 2.516963802850416e-05, + "loss": 0.037, + "step": 40930 + }, + { + "epoch": 1.2636582505092906, + "grad_norm": 0.42506954073905945, + "learning_rate": 2.5145722283064698e-05, + "loss": 0.0381, + "step": 40940 + }, + { + "epoch": 1.2639669115377492, + "grad_norm": 0.48035678267478943, + "learning_rate": 2.5121814087698602e-05, + "loss": 0.0364, + "step": 40950 + }, + { + "epoch": 1.2642755725662078, + "grad_norm": 0.5415387153625488, + "learning_rate": 2.509791344966848e-05, + "loss": 0.0405, + "step": 40960 + }, + { + "epoch": 1.2645842335946664, + "grad_norm": 0.3203616142272949, + "learning_rate": 2.5074020376234768e-05, + "loss": 0.038, + "step": 40970 + }, + { + "epoch": 1.2648928946231248, + "grad_norm": 0.41801345348358154, + "learning_rate": 2.5050134874655534e-05, + "loss": 0.0383, + "step": 40980 + }, + { + "epoch": 1.2652015556515834, + "grad_norm": 0.3113149404525757, + "learning_rate": 2.5026256952186566e-05, + "loss": 0.039, + "step": 40990 + }, + { + "epoch": 1.265510216680042, + "grad_norm": 0.5537382364273071, + "learning_rate": 2.5002386616081335e-05, + "loss": 0.0379, + "step": 41000 + }, + { + "epoch": 1.2658188777085004, + "grad_norm": 0.4753584861755371, + "learning_rate": 2.497852387359103e-05, + "loss": 0.0366, + "step": 41010 + }, + { + "epoch": 1.266127538736959, + "grad_norm": 0.30918335914611816, + "learning_rate": 2.4954668731964496e-05, + "loss": 0.039, + "step": 41020 + }, + { + "epoch": 1.2664361997654177, + "grad_norm": 0.2902798652648926, + "learning_rate": 2.4930821198448364e-05, + "loss": 0.0369, + "step": 41030 + }, + { + "epoch": 1.2667448607938763, + "grad_norm": 0.2489612102508545, + "learning_rate": 2.4906981280286796e-05, + "loss": 0.0362, + "step": 41040 + }, + { + "epoch": 1.2670535218223347, + "grad_norm": 0.2888190746307373, + "learning_rate": 2.488314898472179e-05, + "loss": 0.0372, + "step": 41050 + }, + { + "epoch": 1.2673621828507933, + "grad_norm": 0.5292133092880249, + "learning_rate": 2.485932431899295e-05, + "loss": 0.0355, + "step": 41060 + }, + { + "epoch": 1.2676708438792519, + "grad_norm": 0.3226526081562042, + "learning_rate": 2.4835507290337584e-05, + "loss": 0.0364, + "step": 41070 + }, + { + "epoch": 1.2679795049077103, + "grad_norm": 0.33431851863861084, + "learning_rate": 2.4811697905990672e-05, + "loss": 0.0365, + "step": 41080 + }, + { + "epoch": 1.2682881659361689, + "grad_norm": 0.5073012113571167, + "learning_rate": 2.4787896173184854e-05, + "loss": 0.0377, + "step": 41090 + }, + { + "epoch": 1.2685968269646275, + "grad_norm": 0.3384258449077606, + "learning_rate": 2.4764102099150534e-05, + "loss": 0.0363, + "step": 41100 + }, + { + "epoch": 1.2689054879930861, + "grad_norm": 0.3595525324344635, + "learning_rate": 2.4740315691115644e-05, + "loss": 0.0391, + "step": 41110 + }, + { + "epoch": 1.2692141490215445, + "grad_norm": 0.4470962584018707, + "learning_rate": 2.4716536956305918e-05, + "loss": 0.0382, + "step": 41120 + }, + { + "epoch": 1.2695228100500031, + "grad_norm": 0.384797602891922, + "learning_rate": 2.4692765901944697e-05, + "loss": 0.0365, + "step": 41130 + }, + { + "epoch": 1.2698314710784615, + "grad_norm": 0.2992219924926758, + "learning_rate": 2.4669002535253e-05, + "loss": 0.0375, + "step": 41140 + }, + { + "epoch": 1.2701401321069201, + "grad_norm": 0.4080532193183899, + "learning_rate": 2.46452468634495e-05, + "loss": 0.0374, + "step": 41150 + }, + { + "epoch": 1.2704487931353787, + "grad_norm": 0.36818012595176697, + "learning_rate": 2.462149889375055e-05, + "loss": 0.0371, + "step": 41160 + }, + { + "epoch": 1.2707574541638373, + "grad_norm": 0.39977988600730896, + "learning_rate": 2.459775863337014e-05, + "loss": 0.0369, + "step": 41170 + }, + { + "epoch": 1.271066115192296, + "grad_norm": 0.7948590517044067, + "learning_rate": 2.4574026089519985e-05, + "loss": 0.0434, + "step": 41180 + }, + { + "epoch": 1.2713747762207543, + "grad_norm": 0.5888471603393555, + "learning_rate": 2.4550301269409333e-05, + "loss": 0.0374, + "step": 41190 + }, + { + "epoch": 1.271683437249213, + "grad_norm": 0.41773319244384766, + "learning_rate": 2.4526584180245216e-05, + "loss": 0.0376, + "step": 41200 + }, + { + "epoch": 1.2719920982776713, + "grad_norm": 0.36246728897094727, + "learning_rate": 2.4502874829232236e-05, + "loss": 0.0381, + "step": 41210 + }, + { + "epoch": 1.27230075930613, + "grad_norm": 0.34738674759864807, + "learning_rate": 2.447917322357267e-05, + "loss": 0.037, + "step": 41220 + }, + { + "epoch": 1.2726094203345886, + "grad_norm": 0.2447974979877472, + "learning_rate": 2.4455479370466443e-05, + "loss": 0.0356, + "step": 41230 + }, + { + "epoch": 1.2729180813630472, + "grad_norm": 0.36350902915000916, + "learning_rate": 2.4431793277111097e-05, + "loss": 0.0381, + "step": 41240 + }, + { + "epoch": 1.2732267423915056, + "grad_norm": 0.28055158257484436, + "learning_rate": 2.4408114950701905e-05, + "loss": 0.0388, + "step": 41250 + }, + { + "epoch": 1.2735354034199642, + "grad_norm": 0.3688341975212097, + "learning_rate": 2.4384444398431634e-05, + "loss": 0.035, + "step": 41260 + }, + { + "epoch": 1.2738440644484228, + "grad_norm": 0.3179704546928406, + "learning_rate": 2.4360781627490837e-05, + "loss": 0.0357, + "step": 41270 + }, + { + "epoch": 1.2741527254768812, + "grad_norm": 0.35900670289993286, + "learning_rate": 2.433712664506762e-05, + "loss": 0.0365, + "step": 41280 + }, + { + "epoch": 1.2744613865053398, + "grad_norm": 0.3230055272579193, + "learning_rate": 2.431347945834774e-05, + "loss": 0.0368, + "step": 41290 + }, + { + "epoch": 1.2747700475337984, + "grad_norm": 0.3499939441680908, + "learning_rate": 2.428984007451458e-05, + "loss": 0.0428, + "step": 41300 + }, + { + "epoch": 1.275078708562257, + "grad_norm": 0.26309987902641296, + "learning_rate": 2.426620850074917e-05, + "loss": 0.0358, + "step": 41310 + }, + { + "epoch": 1.2753873695907154, + "grad_norm": 0.2657984793186188, + "learning_rate": 2.424258474423014e-05, + "loss": 0.0393, + "step": 41320 + }, + { + "epoch": 1.275696030619174, + "grad_norm": 0.28074291348457336, + "learning_rate": 2.421896881213382e-05, + "loss": 0.0394, + "step": 41330 + }, + { + "epoch": 1.2760046916476326, + "grad_norm": 0.3725287616252899, + "learning_rate": 2.419536071163402e-05, + "loss": 0.0358, + "step": 41340 + }, + { + "epoch": 1.276313352676091, + "grad_norm": 0.4576760232448578, + "learning_rate": 2.417176044990233e-05, + "loss": 0.0364, + "step": 41350 + }, + { + "epoch": 1.2766220137045496, + "grad_norm": 0.4748617112636566, + "learning_rate": 2.4148168034107855e-05, + "loss": 0.0357, + "step": 41360 + }, + { + "epoch": 1.2769306747330083, + "grad_norm": 0.4103613793849945, + "learning_rate": 2.4124583471417355e-05, + "loss": 0.0391, + "step": 41370 + }, + { + "epoch": 1.2772393357614669, + "grad_norm": 0.38771599531173706, + "learning_rate": 2.41010067689952e-05, + "loss": 0.0389, + "step": 41380 + }, + { + "epoch": 1.2775479967899253, + "grad_norm": 0.32256996631622314, + "learning_rate": 2.4077437934003338e-05, + "loss": 0.0407, + "step": 41390 + }, + { + "epoch": 1.2778566578183839, + "grad_norm": 0.283510684967041, + "learning_rate": 2.405387697360143e-05, + "loss": 0.0402, + "step": 41400 + }, + { + "epoch": 1.2781653188468425, + "grad_norm": 0.30200719833374023, + "learning_rate": 2.4030323894946595e-05, + "loss": 0.0372, + "step": 41410 + }, + { + "epoch": 1.2784739798753009, + "grad_norm": 0.44747084379196167, + "learning_rate": 2.40067787051937e-05, + "loss": 0.0329, + "step": 41420 + }, + { + "epoch": 1.2787826409037595, + "grad_norm": 0.35746192932128906, + "learning_rate": 2.3983241411495087e-05, + "loss": 0.039, + "step": 41430 + }, + { + "epoch": 1.279091301932218, + "grad_norm": 0.427303671836853, + "learning_rate": 2.3959712021000823e-05, + "loss": 0.041, + "step": 41440 + }, + { + "epoch": 1.2793999629606767, + "grad_norm": 0.41589874029159546, + "learning_rate": 2.3936190540858495e-05, + "loss": 0.0385, + "step": 41450 + }, + { + "epoch": 1.279708623989135, + "grad_norm": 0.3785233199596405, + "learning_rate": 2.39126769782133e-05, + "loss": 0.0322, + "step": 41460 + }, + { + "epoch": 1.2800172850175937, + "grad_norm": 0.3408844470977783, + "learning_rate": 2.388917134020805e-05, + "loss": 0.042, + "step": 41470 + }, + { + "epoch": 1.280325946046052, + "grad_norm": 0.289598673582077, + "learning_rate": 2.3865673633983128e-05, + "loss": 0.0369, + "step": 41480 + }, + { + "epoch": 1.2806346070745107, + "grad_norm": 0.2844277322292328, + "learning_rate": 2.3842183866676492e-05, + "loss": 0.0392, + "step": 41490 + }, + { + "epoch": 1.2809432681029693, + "grad_norm": 0.45570021867752075, + "learning_rate": 2.381870204542377e-05, + "loss": 0.035, + "step": 41500 + }, + { + "epoch": 1.281251929131428, + "grad_norm": 0.3246318995952606, + "learning_rate": 2.379522817735808e-05, + "loss": 0.0398, + "step": 41510 + }, + { + "epoch": 1.2815605901598865, + "grad_norm": 0.34819740056991577, + "learning_rate": 2.377176226961018e-05, + "loss": 0.0383, + "step": 41520 + }, + { + "epoch": 1.281869251188345, + "grad_norm": 0.6497821807861328, + "learning_rate": 2.3748304329308384e-05, + "loss": 0.0395, + "step": 41530 + }, + { + "epoch": 1.2821779122168036, + "grad_norm": 0.4281989336013794, + "learning_rate": 2.372485436357858e-05, + "loss": 0.0307, + "step": 41540 + }, + { + "epoch": 1.282486573245262, + "grad_norm": 0.4831574559211731, + "learning_rate": 2.3701412379544296e-05, + "loss": 0.0376, + "step": 41550 + }, + { + "epoch": 1.2827952342737206, + "grad_norm": 0.27921611070632935, + "learning_rate": 2.367797838432653e-05, + "loss": 0.0397, + "step": 41560 + }, + { + "epoch": 1.2831038953021792, + "grad_norm": 0.44905734062194824, + "learning_rate": 2.3654552385043967e-05, + "loss": 0.038, + "step": 41570 + }, + { + "epoch": 1.2834125563306378, + "grad_norm": 0.2486681044101715, + "learning_rate": 2.3631134388812742e-05, + "loss": 0.0385, + "step": 41580 + }, + { + "epoch": 1.2837212173590962, + "grad_norm": 0.35805588960647583, + "learning_rate": 2.3607724402746684e-05, + "loss": 0.0348, + "step": 41590 + }, + { + "epoch": 1.2840298783875548, + "grad_norm": 0.40371447801589966, + "learning_rate": 2.35843224339571e-05, + "loss": 0.0341, + "step": 41600 + }, + { + "epoch": 1.2843385394160134, + "grad_norm": 0.23347561061382294, + "learning_rate": 2.3560928489552897e-05, + "loss": 0.0359, + "step": 41610 + }, + { + "epoch": 1.2846472004444718, + "grad_norm": 0.4528636634349823, + "learning_rate": 2.353754257664053e-05, + "loss": 0.0375, + "step": 41620 + }, + { + "epoch": 1.2849558614729304, + "grad_norm": 0.2887958586215973, + "learning_rate": 2.3514164702324037e-05, + "loss": 0.0396, + "step": 41630 + }, + { + "epoch": 1.285264522501389, + "grad_norm": 0.3867659568786621, + "learning_rate": 2.3490794873704963e-05, + "loss": 0.0377, + "step": 41640 + }, + { + "epoch": 1.2855731835298476, + "grad_norm": 0.42000433802604675, + "learning_rate": 2.3467433097882496e-05, + "loss": 0.0339, + "step": 41650 + }, + { + "epoch": 1.285881844558306, + "grad_norm": 0.2753564417362213, + "learning_rate": 2.34440793819533e-05, + "loss": 0.0393, + "step": 41660 + }, + { + "epoch": 1.2861905055867646, + "grad_norm": 0.3609859049320221, + "learning_rate": 2.3420733733011617e-05, + "loss": 0.0371, + "step": 41670 + }, + { + "epoch": 1.2864991666152232, + "grad_norm": 0.38781091570854187, + "learning_rate": 2.3397396158149243e-05, + "loss": 0.0344, + "step": 41680 + }, + { + "epoch": 1.2868078276436816, + "grad_norm": 0.4328981935977936, + "learning_rate": 2.3374066664455498e-05, + "loss": 0.0371, + "step": 41690 + }, + { + "epoch": 1.2871164886721402, + "grad_norm": 0.2678567171096802, + "learning_rate": 2.3350745259017315e-05, + "loss": 0.0378, + "step": 41700 + }, + { + "epoch": 1.2874251497005988, + "grad_norm": 0.2571996748447418, + "learning_rate": 2.332743194891906e-05, + "loss": 0.0342, + "step": 41710 + }, + { + "epoch": 1.2877338107290575, + "grad_norm": 0.27076882123947144, + "learning_rate": 2.330412674124276e-05, + "loss": 0.0376, + "step": 41720 + }, + { + "epoch": 1.2880424717575158, + "grad_norm": 0.3182317912578583, + "learning_rate": 2.328082964306786e-05, + "loss": 0.0364, + "step": 41730 + }, + { + "epoch": 1.2883511327859745, + "grad_norm": 0.42319628596305847, + "learning_rate": 2.325754066147145e-05, + "loss": 0.0367, + "step": 41740 + }, + { + "epoch": 1.2886597938144329, + "grad_norm": 0.8401144742965698, + "learning_rate": 2.32342598035281e-05, + "loss": 0.0333, + "step": 41750 + }, + { + "epoch": 1.2889684548428915, + "grad_norm": 0.34935274720191956, + "learning_rate": 2.321098707630991e-05, + "loss": 0.0382, + "step": 41760 + }, + { + "epoch": 1.28927711587135, + "grad_norm": 0.5786803364753723, + "learning_rate": 2.318772248688652e-05, + "loss": 0.0366, + "step": 41770 + }, + { + "epoch": 1.2895857768998087, + "grad_norm": 0.2798984944820404, + "learning_rate": 2.3164466042325107e-05, + "loss": 0.0426, + "step": 41780 + }, + { + "epoch": 1.2898944379282673, + "grad_norm": 0.28892940282821655, + "learning_rate": 2.3141217749690353e-05, + "loss": 0.0408, + "step": 41790 + }, + { + "epoch": 1.2902030989567257, + "grad_norm": 0.47075897455215454, + "learning_rate": 2.3117977616044466e-05, + "loss": 0.0365, + "step": 41800 + }, + { + "epoch": 1.2905117599851843, + "grad_norm": 0.22617092728614807, + "learning_rate": 2.309474564844722e-05, + "loss": 0.0315, + "step": 41810 + }, + { + "epoch": 1.2908204210136427, + "grad_norm": 0.3408479392528534, + "learning_rate": 2.307152185395585e-05, + "loss": 0.0376, + "step": 41820 + }, + { + "epoch": 1.2911290820421013, + "grad_norm": 0.38819795846939087, + "learning_rate": 2.3048306239625144e-05, + "loss": 0.0403, + "step": 41830 + }, + { + "epoch": 1.29143774307056, + "grad_norm": 0.35511302947998047, + "learning_rate": 2.3025098812507378e-05, + "loss": 0.0391, + "step": 41840 + }, + { + "epoch": 1.2917464040990185, + "grad_norm": 0.39610257744789124, + "learning_rate": 2.3001899579652366e-05, + "loss": 0.0379, + "step": 41850 + }, + { + "epoch": 1.2920550651274771, + "grad_norm": 0.28097257018089294, + "learning_rate": 2.2978708548107393e-05, + "loss": 0.0344, + "step": 41860 + }, + { + "epoch": 1.2923637261559355, + "grad_norm": 0.32742005586624146, + "learning_rate": 2.2955525724917348e-05, + "loss": 0.0374, + "step": 41870 + }, + { + "epoch": 1.2926723871843941, + "grad_norm": 0.26804351806640625, + "learning_rate": 2.2932351117124477e-05, + "loss": 0.0392, + "step": 41880 + }, + { + "epoch": 1.2929810482128525, + "grad_norm": 0.4391234815120697, + "learning_rate": 2.29091847317687e-05, + "loss": 0.0386, + "step": 41890 + }, + { + "epoch": 1.2932897092413111, + "grad_norm": 0.4681857228279114, + "learning_rate": 2.2886026575887277e-05, + "loss": 0.0406, + "step": 41900 + }, + { + "epoch": 1.2935983702697698, + "grad_norm": 0.3001538813114166, + "learning_rate": 2.2862876656515094e-05, + "loss": 0.037, + "step": 41910 + }, + { + "epoch": 1.2939070312982284, + "grad_norm": 0.23282788693904877, + "learning_rate": 2.2839734980684464e-05, + "loss": 0.0323, + "step": 41920 + }, + { + "epoch": 1.2942156923266868, + "grad_norm": 0.32787957787513733, + "learning_rate": 2.281660155542522e-05, + "loss": 0.0346, + "step": 41930 + }, + { + "epoch": 1.2945243533551454, + "grad_norm": 0.22924809157848358, + "learning_rate": 2.279347638776469e-05, + "loss": 0.039, + "step": 41940 + }, + { + "epoch": 1.294833014383604, + "grad_norm": 0.44494354724884033, + "learning_rate": 2.2770359484727665e-05, + "loss": 0.0382, + "step": 41950 + }, + { + "epoch": 1.2951416754120624, + "grad_norm": 0.43250375986099243, + "learning_rate": 2.27472508533365e-05, + "loss": 0.0379, + "step": 41960 + }, + { + "epoch": 1.295450336440521, + "grad_norm": 0.41718801856040955, + "learning_rate": 2.2724150500610948e-05, + "loss": 0.0397, + "step": 41970 + }, + { + "epoch": 1.2957589974689796, + "grad_norm": 0.2864798605442047, + "learning_rate": 2.2701058433568302e-05, + "loss": 0.0368, + "step": 41980 + }, + { + "epoch": 1.2960676584974382, + "grad_norm": 0.5292531847953796, + "learning_rate": 2.2677974659223318e-05, + "loss": 0.0319, + "step": 41990 + }, + { + "epoch": 1.2963763195258966, + "grad_norm": 0.32070866227149963, + "learning_rate": 2.2654899184588235e-05, + "loss": 0.033, + "step": 42000 + }, + { + "epoch": 1.2966849805543552, + "grad_norm": 0.46960604190826416, + "learning_rate": 2.2631832016672756e-05, + "loss": 0.0389, + "step": 42010 + }, + { + "epoch": 1.2969936415828138, + "grad_norm": 0.37913304567337036, + "learning_rate": 2.2608773162484127e-05, + "loss": 0.0355, + "step": 42020 + }, + { + "epoch": 1.2973023026112722, + "grad_norm": 0.26189860701560974, + "learning_rate": 2.2585722629026958e-05, + "loss": 0.0351, + "step": 42030 + }, + { + "epoch": 1.2976109636397308, + "grad_norm": 0.4114587604999542, + "learning_rate": 2.2562680423303457e-05, + "loss": 0.0368, + "step": 42040 + }, + { + "epoch": 1.2979196246681894, + "grad_norm": 0.29940125346183777, + "learning_rate": 2.2539646552313165e-05, + "loss": 0.0347, + "step": 42050 + }, + { + "epoch": 1.298228285696648, + "grad_norm": 0.35477593541145325, + "learning_rate": 2.251662102305322e-05, + "loss": 0.0372, + "step": 42060 + }, + { + "epoch": 1.2985369467251064, + "grad_norm": 0.47228023409843445, + "learning_rate": 2.2493603842518152e-05, + "loss": 0.0442, + "step": 42070 + }, + { + "epoch": 1.298845607753565, + "grad_norm": 0.35520029067993164, + "learning_rate": 2.2470595017699974e-05, + "loss": 0.0386, + "step": 42080 + }, + { + "epoch": 1.2991542687820234, + "grad_norm": 0.8126820921897888, + "learning_rate": 2.244759455558816e-05, + "loss": 0.0328, + "step": 42090 + }, + { + "epoch": 1.299462929810482, + "grad_norm": 0.27737078070640564, + "learning_rate": 2.2424602463169614e-05, + "loss": 0.0315, + "step": 42100 + }, + { + "epoch": 1.2997715908389407, + "grad_norm": 0.4681381583213806, + "learning_rate": 2.2401618747428776e-05, + "loss": 0.0398, + "step": 42110 + }, + { + "epoch": 1.3000802518673993, + "grad_norm": 0.3065340518951416, + "learning_rate": 2.237864341534747e-05, + "loss": 0.0352, + "step": 42120 + }, + { + "epoch": 1.300388912895858, + "grad_norm": 0.33089473843574524, + "learning_rate": 2.2355676473904998e-05, + "loss": 0.0343, + "step": 42130 + }, + { + "epoch": 1.3006975739243163, + "grad_norm": 0.3815844655036926, + "learning_rate": 2.2332717930078108e-05, + "loss": 0.0373, + "step": 42140 + }, + { + "epoch": 1.301006234952775, + "grad_norm": 0.5240123271942139, + "learning_rate": 2.2309767790840992e-05, + "loss": 0.0384, + "step": 42150 + }, + { + "epoch": 1.3013148959812333, + "grad_norm": 0.2689821124076843, + "learning_rate": 2.228682606316529e-05, + "loss": 0.0379, + "step": 42160 + }, + { + "epoch": 1.301623557009692, + "grad_norm": 0.26437899470329285, + "learning_rate": 2.2263892754020138e-05, + "loss": 0.0375, + "step": 42170 + }, + { + "epoch": 1.3019322180381505, + "grad_norm": 0.30271726846694946, + "learning_rate": 2.2240967870372004e-05, + "loss": 0.036, + "step": 42180 + }, + { + "epoch": 1.3022408790666091, + "grad_norm": 0.45827460289001465, + "learning_rate": 2.2218051419184933e-05, + "loss": 0.0385, + "step": 42190 + }, + { + "epoch": 1.3025495400950675, + "grad_norm": 1.0328457355499268, + "learning_rate": 2.219514340742026e-05, + "loss": 0.0408, + "step": 42200 + }, + { + "epoch": 1.3028582011235261, + "grad_norm": 0.3437698781490326, + "learning_rate": 2.2172243842036898e-05, + "loss": 0.0363, + "step": 42210 + }, + { + "epoch": 1.3031668621519847, + "grad_norm": 0.2785325050354004, + "learning_rate": 2.2149352729991107e-05, + "loss": 0.036, + "step": 42220 + }, + { + "epoch": 1.3034755231804431, + "grad_norm": 0.4718206524848938, + "learning_rate": 2.2126470078236605e-05, + "loss": 0.0375, + "step": 42230 + }, + { + "epoch": 1.3037841842089017, + "grad_norm": 0.24973933398723602, + "learning_rate": 2.2103595893724533e-05, + "loss": 0.0349, + "step": 42240 + }, + { + "epoch": 1.3040928452373604, + "grad_norm": 0.3240271806716919, + "learning_rate": 2.208073018340345e-05, + "loss": 0.035, + "step": 42250 + }, + { + "epoch": 1.304401506265819, + "grad_norm": 0.4958142936229706, + "learning_rate": 2.2057872954219405e-05, + "loss": 0.0435, + "step": 42260 + }, + { + "epoch": 1.3047101672942774, + "grad_norm": 0.32080498337745667, + "learning_rate": 2.203502421311575e-05, + "loss": 0.0419, + "step": 42270 + }, + { + "epoch": 1.305018828322736, + "grad_norm": 0.36041635274887085, + "learning_rate": 2.2012183967033388e-05, + "loss": 0.037, + "step": 42280 + }, + { + "epoch": 1.3053274893511946, + "grad_norm": 0.2529146373271942, + "learning_rate": 2.198935222291056e-05, + "loss": 0.0384, + "step": 42290 + }, + { + "epoch": 1.305636150379653, + "grad_norm": 0.2486579418182373, + "learning_rate": 2.1966528987682948e-05, + "loss": 0.0356, + "step": 42300 + }, + { + "epoch": 1.3059448114081116, + "grad_norm": 0.5821029543876648, + "learning_rate": 2.194371426828365e-05, + "loss": 0.0365, + "step": 42310 + }, + { + "epoch": 1.3062534724365702, + "grad_norm": 0.3201395273208618, + "learning_rate": 2.192090807164317e-05, + "loss": 0.0401, + "step": 42320 + }, + { + "epoch": 1.3065621334650288, + "grad_norm": 0.27880844473838806, + "learning_rate": 2.1898110404689422e-05, + "loss": 0.0392, + "step": 42330 + }, + { + "epoch": 1.3068707944934872, + "grad_norm": 0.36781394481658936, + "learning_rate": 2.1875321274347776e-05, + "loss": 0.0361, + "step": 42340 + }, + { + "epoch": 1.3071794555219458, + "grad_norm": 0.31200823187828064, + "learning_rate": 2.18525406875409e-05, + "loss": 0.0342, + "step": 42350 + }, + { + "epoch": 1.3074881165504044, + "grad_norm": 0.28296902775764465, + "learning_rate": 2.1829768651188997e-05, + "loss": 0.0325, + "step": 42360 + }, + { + "epoch": 1.3077967775788628, + "grad_norm": 0.29550689458847046, + "learning_rate": 2.180700517220958e-05, + "loss": 0.044, + "step": 42370 + }, + { + "epoch": 1.3081054386073214, + "grad_norm": 0.29800674319267273, + "learning_rate": 2.1784250257517603e-05, + "loss": 0.0396, + "step": 42380 + }, + { + "epoch": 1.30841409963578, + "grad_norm": 0.34138020873069763, + "learning_rate": 2.1761503914025406e-05, + "loss": 0.0399, + "step": 42390 + }, + { + "epoch": 1.3087227606642386, + "grad_norm": 0.36674487590789795, + "learning_rate": 2.1738766148642705e-05, + "loss": 0.0342, + "step": 42400 + }, + { + "epoch": 1.309031421692697, + "grad_norm": 0.3505118191242218, + "learning_rate": 2.1716036968276683e-05, + "loss": 0.0398, + "step": 42410 + }, + { + "epoch": 1.3093400827211557, + "grad_norm": 0.3177063763141632, + "learning_rate": 2.1693316379831808e-05, + "loss": 0.0426, + "step": 42420 + }, + { + "epoch": 1.309648743749614, + "grad_norm": 0.29065948724746704, + "learning_rate": 2.1670604390210037e-05, + "loss": 0.0374, + "step": 42430 + }, + { + "epoch": 1.3099574047780727, + "grad_norm": 0.4786668121814728, + "learning_rate": 2.1647901006310656e-05, + "loss": 0.0412, + "step": 42440 + }, + { + "epoch": 1.3102660658065313, + "grad_norm": 0.42397525906562805, + "learning_rate": 2.1625206235030353e-05, + "loss": 0.0354, + "step": 42450 + }, + { + "epoch": 1.3105747268349899, + "grad_norm": 0.3754245936870575, + "learning_rate": 2.160252008326321e-05, + "loss": 0.0342, + "step": 42460 + }, + { + "epoch": 1.3108833878634485, + "grad_norm": 0.2936563193798065, + "learning_rate": 2.157984255790067e-05, + "loss": 0.041, + "step": 42470 + }, + { + "epoch": 1.3111920488919069, + "grad_norm": 0.5891686081886292, + "learning_rate": 2.1557173665831553e-05, + "loss": 0.0365, + "step": 42480 + }, + { + "epoch": 1.3115007099203655, + "grad_norm": 0.3544497787952423, + "learning_rate": 2.153451341394212e-05, + "loss": 0.0385, + "step": 42490 + }, + { + "epoch": 1.3118093709488239, + "grad_norm": 0.4182102084159851, + "learning_rate": 2.151186180911589e-05, + "loss": 0.0361, + "step": 42500 + }, + { + "epoch": 1.3121180319772825, + "grad_norm": 0.3503898084163666, + "learning_rate": 2.1489218858233877e-05, + "loss": 0.0353, + "step": 42510 + }, + { + "epoch": 1.312426693005741, + "grad_norm": 0.7096041440963745, + "learning_rate": 2.1466584568174392e-05, + "loss": 0.0394, + "step": 42520 + }, + { + "epoch": 1.3127353540341997, + "grad_norm": 0.31642359495162964, + "learning_rate": 2.1443958945813132e-05, + "loss": 0.0345, + "step": 42530 + }, + { + "epoch": 1.313044015062658, + "grad_norm": 0.375180721282959, + "learning_rate": 2.1421341998023163e-05, + "loss": 0.0419, + "step": 42540 + }, + { + "epoch": 1.3133526760911167, + "grad_norm": 0.3376173973083496, + "learning_rate": 2.139873373167491e-05, + "loss": 0.0378, + "step": 42550 + }, + { + "epoch": 1.3136613371195753, + "grad_norm": 0.30801481008529663, + "learning_rate": 2.13761341536362e-05, + "loss": 0.0385, + "step": 42560 + }, + { + "epoch": 1.3139699981480337, + "grad_norm": 0.34591153264045715, + "learning_rate": 2.1353543270772136e-05, + "loss": 0.0376, + "step": 42570 + }, + { + "epoch": 1.3142786591764923, + "grad_norm": 0.41731229424476624, + "learning_rate": 2.1330961089945297e-05, + "loss": 0.0375, + "step": 42580 + }, + { + "epoch": 1.314587320204951, + "grad_norm": 0.2704162299633026, + "learning_rate": 2.130838761801548e-05, + "loss": 0.037, + "step": 42590 + }, + { + "epoch": 1.3148959812334096, + "grad_norm": 0.3986648917198181, + "learning_rate": 2.1285822861839966e-05, + "loss": 0.0383, + "step": 42600 + }, + { + "epoch": 1.315204642261868, + "grad_norm": 0.2531539499759674, + "learning_rate": 2.126326682827331e-05, + "loss": 0.0368, + "step": 42610 + }, + { + "epoch": 1.3155133032903266, + "grad_norm": 0.3090081810951233, + "learning_rate": 2.124071952416744e-05, + "loss": 0.0353, + "step": 42620 + }, + { + "epoch": 1.3158219643187852, + "grad_norm": 0.3209497630596161, + "learning_rate": 2.1218180956371634e-05, + "loss": 0.0375, + "step": 42630 + }, + { + "epoch": 1.3161306253472436, + "grad_norm": 0.3536710739135742, + "learning_rate": 2.119565113173252e-05, + "loss": 0.038, + "step": 42640 + }, + { + "epoch": 1.3164392863757022, + "grad_norm": 0.30686911940574646, + "learning_rate": 2.1173130057094033e-05, + "loss": 0.0333, + "step": 42650 + }, + { + "epoch": 1.3167479474041608, + "grad_norm": 0.2837987244129181, + "learning_rate": 2.115061773929753e-05, + "loss": 0.037, + "step": 42660 + }, + { + "epoch": 1.3170566084326194, + "grad_norm": 0.23370860517024994, + "learning_rate": 2.1128114185181623e-05, + "loss": 0.0403, + "step": 42670 + }, + { + "epoch": 1.3173652694610778, + "grad_norm": 0.35146406292915344, + "learning_rate": 2.1105619401582317e-05, + "loss": 0.0361, + "step": 42680 + }, + { + "epoch": 1.3176739304895364, + "grad_norm": 0.31140127778053284, + "learning_rate": 2.1083133395332928e-05, + "loss": 0.0346, + "step": 42690 + }, + { + "epoch": 1.317982591517995, + "grad_norm": 0.20837225019931793, + "learning_rate": 2.1060656173264082e-05, + "loss": 0.0371, + "step": 42700 + }, + { + "epoch": 1.3182912525464534, + "grad_norm": 0.2547198534011841, + "learning_rate": 2.103818774220383e-05, + "loss": 0.0357, + "step": 42710 + }, + { + "epoch": 1.318599913574912, + "grad_norm": 0.34156179428100586, + "learning_rate": 2.1015728108977412e-05, + "loss": 0.0362, + "step": 42720 + }, + { + "epoch": 1.3189085746033706, + "grad_norm": 0.3049685060977936, + "learning_rate": 2.0993277280407548e-05, + "loss": 0.0379, + "step": 42730 + }, + { + "epoch": 1.3192172356318292, + "grad_norm": 0.47965824604034424, + "learning_rate": 2.0970835263314132e-05, + "loss": 0.0357, + "step": 42740 + }, + { + "epoch": 1.3195258966602876, + "grad_norm": 0.2930842339992523, + "learning_rate": 2.094840206451451e-05, + "loss": 0.0365, + "step": 42750 + }, + { + "epoch": 1.3198345576887462, + "grad_norm": 0.2550834119319916, + "learning_rate": 2.0925977690823273e-05, + "loss": 0.0309, + "step": 42760 + }, + { + "epoch": 1.3201432187172046, + "grad_norm": 0.4306902587413788, + "learning_rate": 2.0903562149052364e-05, + "loss": 0.0358, + "step": 42770 + }, + { + "epoch": 1.3204518797456632, + "grad_norm": 0.3111358880996704, + "learning_rate": 2.0881155446011025e-05, + "loss": 0.0347, + "step": 42780 + }, + { + "epoch": 1.3207605407741219, + "grad_norm": 0.38181981444358826, + "learning_rate": 2.0858757588505823e-05, + "loss": 0.0393, + "step": 42790 + }, + { + "epoch": 1.3210692018025805, + "grad_norm": 0.43320080637931824, + "learning_rate": 2.0836368583340622e-05, + "loss": 0.0359, + "step": 42800 + }, + { + "epoch": 1.321377862831039, + "grad_norm": 0.3051401674747467, + "learning_rate": 2.081398843731664e-05, + "loss": 0.0367, + "step": 42810 + }, + { + "epoch": 1.3216865238594975, + "grad_norm": 0.4248441159725189, + "learning_rate": 2.0791617157232357e-05, + "loss": 0.0375, + "step": 42820 + }, + { + "epoch": 1.321995184887956, + "grad_norm": 0.2418946623802185, + "learning_rate": 2.0769254749883576e-05, + "loss": 0.0335, + "step": 42830 + }, + { + "epoch": 1.3223038459164145, + "grad_norm": 0.3294658064842224, + "learning_rate": 2.0746901222063415e-05, + "loss": 0.0367, + "step": 42840 + }, + { + "epoch": 1.322612506944873, + "grad_norm": 0.30224448442459106, + "learning_rate": 2.072455658056226e-05, + "loss": 0.0374, + "step": 42850 + }, + { + "epoch": 1.3229211679733317, + "grad_norm": 0.33899134397506714, + "learning_rate": 2.0702220832167873e-05, + "loss": 0.0351, + "step": 42860 + }, + { + "epoch": 1.3232298290017903, + "grad_norm": 0.527771532535553, + "learning_rate": 2.0679893983665205e-05, + "loss": 0.0381, + "step": 42870 + }, + { + "epoch": 1.3235384900302487, + "grad_norm": 0.32551679015159607, + "learning_rate": 2.0657576041836622e-05, + "loss": 0.0388, + "step": 42880 + }, + { + "epoch": 1.3238471510587073, + "grad_norm": 0.36459487676620483, + "learning_rate": 2.0635267013461666e-05, + "loss": 0.0344, + "step": 42890 + }, + { + "epoch": 1.324155812087166, + "grad_norm": 0.46896520256996155, + "learning_rate": 2.061296690531728e-05, + "loss": 0.0399, + "step": 42900 + }, + { + "epoch": 1.3244644731156243, + "grad_norm": 0.43486446142196655, + "learning_rate": 2.0590675724177622e-05, + "loss": 0.0356, + "step": 42910 + }, + { + "epoch": 1.324773134144083, + "grad_norm": 0.3588463068008423, + "learning_rate": 2.0568393476814167e-05, + "loss": 0.0371, + "step": 42920 + }, + { + "epoch": 1.3250817951725415, + "grad_norm": 0.3911956548690796, + "learning_rate": 2.0546120169995685e-05, + "loss": 0.032, + "step": 42930 + }, + { + "epoch": 1.3253904562010002, + "grad_norm": 0.27141448855400085, + "learning_rate": 2.0523855810488214e-05, + "loss": 0.0358, + "step": 42940 + }, + { + "epoch": 1.3256991172294585, + "grad_norm": 0.37077540159225464, + "learning_rate": 2.050160040505505e-05, + "loss": 0.0369, + "step": 42950 + }, + { + "epoch": 1.3260077782579172, + "grad_norm": 0.3008142411708832, + "learning_rate": 2.0479353960456843e-05, + "loss": 0.0314, + "step": 42960 + }, + { + "epoch": 1.3263164392863758, + "grad_norm": 0.3157072365283966, + "learning_rate": 2.0457116483451456e-05, + "loss": 0.0394, + "step": 42970 + }, + { + "epoch": 1.3266251003148342, + "grad_norm": 0.3334549069404602, + "learning_rate": 2.0434887980794043e-05, + "loss": 0.0353, + "step": 42980 + }, + { + "epoch": 1.3269337613432928, + "grad_norm": 0.41436266899108887, + "learning_rate": 2.0412668459237043e-05, + "loss": 0.0352, + "step": 42990 + }, + { + "epoch": 1.3272424223717514, + "grad_norm": 0.2835777699947357, + "learning_rate": 2.039045792553016e-05, + "loss": 0.0371, + "step": 43000 + }, + { + "epoch": 1.32755108340021, + "grad_norm": 0.35088634490966797, + "learning_rate": 2.036825638642036e-05, + "loss": 0.0366, + "step": 43010 + }, + { + "epoch": 1.3278597444286684, + "grad_norm": 0.34623289108276367, + "learning_rate": 2.0346063848651868e-05, + "loss": 0.0425, + "step": 43020 + }, + { + "epoch": 1.328168405457127, + "grad_norm": 0.4527813196182251, + "learning_rate": 2.0323880318966254e-05, + "loss": 0.0378, + "step": 43030 + }, + { + "epoch": 1.3284770664855856, + "grad_norm": 0.4086461067199707, + "learning_rate": 2.030170580410221e-05, + "loss": 0.0389, + "step": 43040 + }, + { + "epoch": 1.328785727514044, + "grad_norm": 0.42764583230018616, + "learning_rate": 2.0279540310795837e-05, + "loss": 0.0367, + "step": 43050 + }, + { + "epoch": 1.3290943885425026, + "grad_norm": 0.48989778757095337, + "learning_rate": 2.0257383845780365e-05, + "loss": 0.0385, + "step": 43060 + }, + { + "epoch": 1.3294030495709612, + "grad_norm": 0.31324684619903564, + "learning_rate": 2.0235236415786384e-05, + "loss": 0.0418, + "step": 43070 + }, + { + "epoch": 1.3297117105994198, + "grad_norm": 0.35882577300071716, + "learning_rate": 2.021309802754169e-05, + "loss": 0.0363, + "step": 43080 + }, + { + "epoch": 1.3300203716278782, + "grad_norm": 0.3317223787307739, + "learning_rate": 2.0190968687771332e-05, + "loss": 0.0369, + "step": 43090 + }, + { + "epoch": 1.3303290326563368, + "grad_norm": 0.27228158712387085, + "learning_rate": 2.016884840319763e-05, + "loss": 0.0391, + "step": 43100 + }, + { + "epoch": 1.3306376936847952, + "grad_norm": 0.3526616394519806, + "learning_rate": 2.0146737180540122e-05, + "loss": 0.0317, + "step": 43110 + }, + { + "epoch": 1.3309463547132538, + "grad_norm": 0.35707011818885803, + "learning_rate": 2.012463502651564e-05, + "loss": 0.0377, + "step": 43120 + }, + { + "epoch": 1.3312550157417125, + "grad_norm": 0.37416893243789673, + "learning_rate": 2.0102541947838228e-05, + "loss": 0.0381, + "step": 43130 + }, + { + "epoch": 1.331563676770171, + "grad_norm": 0.40557512640953064, + "learning_rate": 2.0080457951219173e-05, + "loss": 0.0377, + "step": 43140 + }, + { + "epoch": 1.3318723377986297, + "grad_norm": 0.4453551769256592, + "learning_rate": 2.0058383043367017e-05, + "loss": 0.0366, + "step": 43150 + }, + { + "epoch": 1.332180998827088, + "grad_norm": 0.41633984446525574, + "learning_rate": 2.0036317230987528e-05, + "loss": 0.0348, + "step": 43160 + }, + { + "epoch": 1.3324896598555467, + "grad_norm": 0.48627743124961853, + "learning_rate": 2.0014260520783696e-05, + "loss": 0.0379, + "step": 43170 + }, + { + "epoch": 1.332798320884005, + "grad_norm": 0.3640210032463074, + "learning_rate": 1.9992212919455834e-05, + "loss": 0.037, + "step": 43180 + }, + { + "epoch": 1.3331069819124637, + "grad_norm": 0.32294514775276184, + "learning_rate": 1.9970174433701333e-05, + "loss": 0.036, + "step": 43190 + }, + { + "epoch": 1.3334156429409223, + "grad_norm": 0.27135583758354187, + "learning_rate": 1.9948145070214992e-05, + "loss": 0.0407, + "step": 43200 + }, + { + "epoch": 1.333724303969381, + "grad_norm": 0.3521381616592407, + "learning_rate": 1.9926124835688663e-05, + "loss": 0.0318, + "step": 43210 + }, + { + "epoch": 1.3340329649978393, + "grad_norm": 0.3792300224304199, + "learning_rate": 1.9904113736811576e-05, + "loss": 0.0424, + "step": 43220 + }, + { + "epoch": 1.334341626026298, + "grad_norm": 0.419255793094635, + "learning_rate": 1.9882111780270096e-05, + "loss": 0.0404, + "step": 43230 + }, + { + "epoch": 1.3346502870547565, + "grad_norm": 0.46164384484291077, + "learning_rate": 1.986011897274784e-05, + "loss": 0.0369, + "step": 43240 + }, + { + "epoch": 1.334958948083215, + "grad_norm": 0.3161945939064026, + "learning_rate": 1.983813532092565e-05, + "loss": 0.0357, + "step": 43250 + }, + { + "epoch": 1.3352676091116735, + "grad_norm": 0.4634133279323578, + "learning_rate": 1.981616083148155e-05, + "loss": 0.0396, + "step": 43260 + }, + { + "epoch": 1.3355762701401321, + "grad_norm": 0.2986149191856384, + "learning_rate": 1.9794195511090845e-05, + "loss": 0.0402, + "step": 43270 + }, + { + "epoch": 1.3358849311685908, + "grad_norm": 0.4031067192554474, + "learning_rate": 1.977223936642601e-05, + "loss": 0.0378, + "step": 43280 + }, + { + "epoch": 1.3361935921970491, + "grad_norm": 0.34306836128234863, + "learning_rate": 1.975029240415674e-05, + "loss": 0.0345, + "step": 43290 + }, + { + "epoch": 1.3365022532255078, + "grad_norm": 0.40489739179611206, + "learning_rate": 1.9728354630949936e-05, + "loss": 0.0352, + "step": 43300 + }, + { + "epoch": 1.3368109142539664, + "grad_norm": 0.30234360694885254, + "learning_rate": 1.9706426053469716e-05, + "loss": 0.0371, + "step": 43310 + }, + { + "epoch": 1.3371195752824248, + "grad_norm": 0.311169296503067, + "learning_rate": 1.9684506678377396e-05, + "loss": 0.0398, + "step": 43320 + }, + { + "epoch": 1.3374282363108834, + "grad_norm": 0.28502437472343445, + "learning_rate": 1.9662596512331544e-05, + "loss": 0.0411, + "step": 43330 + }, + { + "epoch": 1.337736897339342, + "grad_norm": 0.4394780695438385, + "learning_rate": 1.964069556198782e-05, + "loss": 0.0362, + "step": 43340 + }, + { + "epoch": 1.3380455583678006, + "grad_norm": 0.3613692820072174, + "learning_rate": 1.9618803833999232e-05, + "loss": 0.0361, + "step": 43350 + }, + { + "epoch": 1.338354219396259, + "grad_norm": 0.4864151179790497, + "learning_rate": 1.9596921335015838e-05, + "loss": 0.0343, + "step": 43360 + }, + { + "epoch": 1.3386628804247176, + "grad_norm": 0.35251834988594055, + "learning_rate": 1.957504807168501e-05, + "loss": 0.0373, + "step": 43370 + }, + { + "epoch": 1.338971541453176, + "grad_norm": 0.28363391757011414, + "learning_rate": 1.9553184050651253e-05, + "loss": 0.0405, + "step": 43380 + }, + { + "epoch": 1.3392802024816346, + "grad_norm": 0.3335096836090088, + "learning_rate": 1.953132927855628e-05, + "loss": 0.0382, + "step": 43390 + }, + { + "epoch": 1.3395888635100932, + "grad_norm": 0.2967945635318756, + "learning_rate": 1.9509483762038995e-05, + "loss": 0.034, + "step": 43400 + }, + { + "epoch": 1.3398975245385518, + "grad_norm": 0.41022956371307373, + "learning_rate": 1.9487647507735467e-05, + "loss": 0.0374, + "step": 43410 + }, + { + "epoch": 1.3402061855670104, + "grad_norm": 0.3404339849948883, + "learning_rate": 1.9465820522279032e-05, + "loss": 0.0331, + "step": 43420 + }, + { + "epoch": 1.3405148465954688, + "grad_norm": 0.5380638837814331, + "learning_rate": 1.9444002812300078e-05, + "loss": 0.0353, + "step": 43430 + }, + { + "epoch": 1.3408235076239274, + "grad_norm": 0.3339303135871887, + "learning_rate": 1.94221943844263e-05, + "loss": 0.0352, + "step": 43440 + }, + { + "epoch": 1.3411321686523858, + "grad_norm": 0.2687798738479614, + "learning_rate": 1.9400395245282515e-05, + "loss": 0.0341, + "step": 43450 + }, + { + "epoch": 1.3414408296808444, + "grad_norm": 0.29844650626182556, + "learning_rate": 1.937860540149071e-05, + "loss": 0.0361, + "step": 43460 + }, + { + "epoch": 1.341749490709303, + "grad_norm": 0.31921112537384033, + "learning_rate": 1.9356824859670082e-05, + "loss": 0.0343, + "step": 43470 + }, + { + "epoch": 1.3420581517377617, + "grad_norm": 0.49509304761886597, + "learning_rate": 1.9335053626436967e-05, + "loss": 0.0354, + "step": 43480 + }, + { + "epoch": 1.3423668127662203, + "grad_norm": 0.28810176253318787, + "learning_rate": 1.9313291708404885e-05, + "loss": 0.0356, + "step": 43490 + }, + { + "epoch": 1.3426754737946787, + "grad_norm": 0.3126058578491211, + "learning_rate": 1.9291539112184587e-05, + "loss": 0.0354, + "step": 43500 + }, + { + "epoch": 1.3429841348231373, + "grad_norm": 0.3237307071685791, + "learning_rate": 1.9269795844383854e-05, + "loss": 0.0413, + "step": 43510 + }, + { + "epoch": 1.3432927958515957, + "grad_norm": 0.4127575159072876, + "learning_rate": 1.9248061911607777e-05, + "loss": 0.036, + "step": 43520 + }, + { + "epoch": 1.3436014568800543, + "grad_norm": 0.33450478315353394, + "learning_rate": 1.9226337320458538e-05, + "loss": 0.0371, + "step": 43530 + }, + { + "epoch": 1.343910117908513, + "grad_norm": 0.25493815541267395, + "learning_rate": 1.9204622077535488e-05, + "loss": 0.0374, + "step": 43540 + }, + { + "epoch": 1.3442187789369715, + "grad_norm": 0.35667160153388977, + "learning_rate": 1.9182916189435147e-05, + "loss": 0.0399, + "step": 43550 + }, + { + "epoch": 1.34452743996543, + "grad_norm": 0.31795889139175415, + "learning_rate": 1.916121966275117e-05, + "loss": 0.0372, + "step": 43560 + }, + { + "epoch": 1.3448361009938885, + "grad_norm": 0.2959121763706207, + "learning_rate": 1.9139532504074443e-05, + "loss": 0.0355, + "step": 43570 + }, + { + "epoch": 1.3451447620223471, + "grad_norm": 0.2583472430706024, + "learning_rate": 1.9117854719992885e-05, + "loss": 0.0351, + "step": 43580 + }, + { + "epoch": 1.3454534230508055, + "grad_norm": 0.3195803165435791, + "learning_rate": 1.9096186317091687e-05, + "loss": 0.041, + "step": 43590 + }, + { + "epoch": 1.3457620840792641, + "grad_norm": 0.47963425517082214, + "learning_rate": 1.9074527301953116e-05, + "loss": 0.0397, + "step": 43600 + }, + { + "epoch": 1.3460707451077227, + "grad_norm": 0.27099671959877014, + "learning_rate": 1.9052877681156607e-05, + "loss": 0.0376, + "step": 43610 + }, + { + "epoch": 1.3463794061361813, + "grad_norm": 0.25640439987182617, + "learning_rate": 1.903123746127875e-05, + "loss": 0.0368, + "step": 43620 + }, + { + "epoch": 1.3466880671646397, + "grad_norm": 0.3034132122993469, + "learning_rate": 1.900960664889327e-05, + "loss": 0.0406, + "step": 43630 + }, + { + "epoch": 1.3469967281930983, + "grad_norm": 0.27810028195381165, + "learning_rate": 1.8987985250571015e-05, + "loss": 0.0357, + "step": 43640 + }, + { + "epoch": 1.347305389221557, + "grad_norm": 0.3120593726634979, + "learning_rate": 1.8966373272880054e-05, + "loss": 0.0389, + "step": 43650 + }, + { + "epoch": 1.3476140502500153, + "grad_norm": 0.39726945757865906, + "learning_rate": 1.8944770722385462e-05, + "loss": 0.038, + "step": 43660 + }, + { + "epoch": 1.347922711278474, + "grad_norm": 0.4501357972621918, + "learning_rate": 1.8923177605649576e-05, + "loss": 0.0347, + "step": 43670 + }, + { + "epoch": 1.3482313723069326, + "grad_norm": 0.2552148401737213, + "learning_rate": 1.8901593929231802e-05, + "loss": 0.0397, + "step": 43680 + }, + { + "epoch": 1.3485400333353912, + "grad_norm": 0.5345863103866577, + "learning_rate": 1.8880019699688684e-05, + "loss": 0.0407, + "step": 43690 + }, + { + "epoch": 1.3488486943638496, + "grad_norm": 0.3686114251613617, + "learning_rate": 1.8858454923573904e-05, + "loss": 0.0403, + "step": 43700 + }, + { + "epoch": 1.3491573553923082, + "grad_norm": 0.3096786439418793, + "learning_rate": 1.8836899607438253e-05, + "loss": 0.0384, + "step": 43710 + }, + { + "epoch": 1.3494660164207666, + "grad_norm": 0.4140005111694336, + "learning_rate": 1.8815353757829723e-05, + "loss": 0.0354, + "step": 43720 + }, + { + "epoch": 1.3497746774492252, + "grad_norm": 0.3556874394416809, + "learning_rate": 1.879381738129331e-05, + "loss": 0.0365, + "step": 43730 + }, + { + "epoch": 1.3500833384776838, + "grad_norm": 0.39910435676574707, + "learning_rate": 1.8772290484371236e-05, + "loss": 0.0341, + "step": 43740 + }, + { + "epoch": 1.3503919995061424, + "grad_norm": 0.2775637209415436, + "learning_rate": 1.8750773073602795e-05, + "loss": 0.0356, + "step": 43750 + }, + { + "epoch": 1.350700660534601, + "grad_norm": 0.45275363326072693, + "learning_rate": 1.8729265155524405e-05, + "loss": 0.0412, + "step": 43760 + }, + { + "epoch": 1.3510093215630594, + "grad_norm": 0.4221143126487732, + "learning_rate": 1.8707766736669607e-05, + "loss": 0.0393, + "step": 43770 + }, + { + "epoch": 1.351317982591518, + "grad_norm": 0.22089295089244843, + "learning_rate": 1.8686277823569055e-05, + "loss": 0.0379, + "step": 43780 + }, + { + "epoch": 1.3516266436199764, + "grad_norm": 0.27333304286003113, + "learning_rate": 1.8664798422750484e-05, + "loss": 0.0355, + "step": 43790 + }, + { + "epoch": 1.351935304648435, + "grad_norm": 0.2675398886203766, + "learning_rate": 1.8643328540738832e-05, + "loss": 0.0445, + "step": 43800 + }, + { + "epoch": 1.3522439656768936, + "grad_norm": 0.34471434354782104, + "learning_rate": 1.862186818405601e-05, + "loss": 0.0362, + "step": 43810 + }, + { + "epoch": 1.3525526267053523, + "grad_norm": 0.26194095611572266, + "learning_rate": 1.8600417359221156e-05, + "loss": 0.0394, + "step": 43820 + }, + { + "epoch": 1.3528612877338106, + "grad_norm": 0.2388538122177124, + "learning_rate": 1.8578976072750454e-05, + "loss": 0.0357, + "step": 43830 + }, + { + "epoch": 1.3531699487622693, + "grad_norm": 0.2931927740573883, + "learning_rate": 1.8557544331157194e-05, + "loss": 0.0389, + "step": 43840 + }, + { + "epoch": 1.3534786097907279, + "grad_norm": 0.38003331422805786, + "learning_rate": 1.8536122140951785e-05, + "loss": 0.0407, + "step": 43850 + }, + { + "epoch": 1.3537872708191863, + "grad_norm": 0.32820481061935425, + "learning_rate": 1.8514709508641688e-05, + "loss": 0.0382, + "step": 43860 + }, + { + "epoch": 1.3540959318476449, + "grad_norm": 0.35820654034614563, + "learning_rate": 1.8493306440731555e-05, + "loss": 0.0377, + "step": 43870 + }, + { + "epoch": 1.3544045928761035, + "grad_norm": 0.3776342272758484, + "learning_rate": 1.8471912943723013e-05, + "loss": 0.0359, + "step": 43880 + }, + { + "epoch": 1.354713253904562, + "grad_norm": 0.2841205596923828, + "learning_rate": 1.8450529024114894e-05, + "loss": 0.0386, + "step": 43890 + }, + { + "epoch": 1.3550219149330205, + "grad_norm": 0.4941132068634033, + "learning_rate": 1.842915468840301e-05, + "loss": 0.0373, + "step": 43900 + }, + { + "epoch": 1.355330575961479, + "grad_norm": 0.3697129487991333, + "learning_rate": 1.840778994308037e-05, + "loss": 0.0415, + "step": 43910 + }, + { + "epoch": 1.3556392369899377, + "grad_norm": 0.295690655708313, + "learning_rate": 1.8386434794637004e-05, + "loss": 0.0367, + "step": 43920 + }, + { + "epoch": 1.355947898018396, + "grad_norm": 0.41995570063591003, + "learning_rate": 1.8365089249560034e-05, + "loss": 0.0384, + "step": 43930 + }, + { + "epoch": 1.3562565590468547, + "grad_norm": 0.2832760214805603, + "learning_rate": 1.8343753314333683e-05, + "loss": 0.0359, + "step": 43940 + }, + { + "epoch": 1.3565652200753133, + "grad_norm": 0.3826022446155548, + "learning_rate": 1.8322426995439236e-05, + "loss": 0.0374, + "step": 43950 + }, + { + "epoch": 1.356873881103772, + "grad_norm": 0.3088719844818115, + "learning_rate": 1.8301110299355058e-05, + "loss": 0.0365, + "step": 43960 + }, + { + "epoch": 1.3571825421322303, + "grad_norm": 0.2695131301879883, + "learning_rate": 1.8279803232556625e-05, + "loss": 0.0373, + "step": 43970 + }, + { + "epoch": 1.357491203160689, + "grad_norm": 0.3461492359638214, + "learning_rate": 1.8258505801516444e-05, + "loss": 0.0413, + "step": 43980 + }, + { + "epoch": 1.3577998641891476, + "grad_norm": 0.35267066955566406, + "learning_rate": 1.8237218012704117e-05, + "loss": 0.0379, + "step": 43990 + }, + { + "epoch": 1.358108525217606, + "grad_norm": 0.4003359377384186, + "learning_rate": 1.821593987258631e-05, + "loss": 0.0359, + "step": 44000 + }, + { + "epoch": 1.3584171862460646, + "grad_norm": 0.34362244606018066, + "learning_rate": 1.8194671387626744e-05, + "loss": 0.0329, + "step": 44010 + }, + { + "epoch": 1.3587258472745232, + "grad_norm": 0.43154066801071167, + "learning_rate": 1.8173412564286276e-05, + "loss": 0.0349, + "step": 44020 + }, + { + "epoch": 1.3590345083029818, + "grad_norm": 0.29506033658981323, + "learning_rate": 1.8152163409022697e-05, + "loss": 0.0354, + "step": 44030 + }, + { + "epoch": 1.3593431693314402, + "grad_norm": 0.4275067448616028, + "learning_rate": 1.8130923928291023e-05, + "loss": 0.0375, + "step": 44040 + }, + { + "epoch": 1.3596518303598988, + "grad_norm": 0.33969807624816895, + "learning_rate": 1.8109694128543163e-05, + "loss": 0.036, + "step": 44050 + }, + { + "epoch": 1.3599604913883572, + "grad_norm": 0.2138861119747162, + "learning_rate": 1.8088474016228237e-05, + "loss": 0.0352, + "step": 44060 + }, + { + "epoch": 1.3602691524168158, + "grad_norm": 0.4665738344192505, + "learning_rate": 1.8067263597792328e-05, + "loss": 0.037, + "step": 44070 + }, + { + "epoch": 1.3605778134452744, + "grad_norm": 0.30098626017570496, + "learning_rate": 1.80460628796786e-05, + "loss": 0.0339, + "step": 44080 + }, + { + "epoch": 1.360886474473733, + "grad_norm": 0.3586289584636688, + "learning_rate": 1.8024871868327276e-05, + "loss": 0.0385, + "step": 44090 + }, + { + "epoch": 1.3611951355021916, + "grad_norm": 0.36487746238708496, + "learning_rate": 1.8003690570175608e-05, + "loss": 0.035, + "step": 44100 + }, + { + "epoch": 1.36150379653065, + "grad_norm": 0.4162101149559021, + "learning_rate": 1.7982518991657943e-05, + "loss": 0.04, + "step": 44110 + }, + { + "epoch": 1.3618124575591086, + "grad_norm": 0.5845028162002563, + "learning_rate": 1.7961357139205643e-05, + "loss": 0.0366, + "step": 44120 + }, + { + "epoch": 1.362121118587567, + "grad_norm": 2.3101515769958496, + "learning_rate": 1.7940205019247108e-05, + "loss": 0.0365, + "step": 44130 + }, + { + "epoch": 1.3624297796160256, + "grad_norm": 0.3407967686653137, + "learning_rate": 1.79190626382078e-05, + "loss": 0.0353, + "step": 44140 + }, + { + "epoch": 1.3627384406444842, + "grad_norm": 0.2851487994194031, + "learning_rate": 1.7897930002510215e-05, + "loss": 0.041, + "step": 44150 + }, + { + "epoch": 1.3630471016729429, + "grad_norm": 0.408404678106308, + "learning_rate": 1.787680711857387e-05, + "loss": 0.0396, + "step": 44160 + }, + { + "epoch": 1.3633557627014012, + "grad_norm": 0.2841833233833313, + "learning_rate": 1.7855693992815398e-05, + "loss": 0.0372, + "step": 44170 + }, + { + "epoch": 1.3636644237298599, + "grad_norm": 0.5260446071624756, + "learning_rate": 1.7834590631648328e-05, + "loss": 0.0384, + "step": 44180 + }, + { + "epoch": 1.3639730847583185, + "grad_norm": 0.5278587937355042, + "learning_rate": 1.7813497041483384e-05, + "loss": 0.0325, + "step": 44190 + }, + { + "epoch": 1.3642817457867769, + "grad_norm": 0.2902243137359619, + "learning_rate": 1.779241322872817e-05, + "loss": 0.0322, + "step": 44200 + }, + { + "epoch": 1.3645904068152355, + "grad_norm": 0.34213414788246155, + "learning_rate": 1.777133919978744e-05, + "loss": 0.0359, + "step": 44210 + }, + { + "epoch": 1.364899067843694, + "grad_norm": 0.3002396523952484, + "learning_rate": 1.7750274961062912e-05, + "loss": 0.0378, + "step": 44220 + }, + { + "epoch": 1.3652077288721527, + "grad_norm": 0.3617195785045624, + "learning_rate": 1.772922051895335e-05, + "loss": 0.0363, + "step": 44230 + }, + { + "epoch": 1.365516389900611, + "grad_norm": 0.3895461857318878, + "learning_rate": 1.770817587985453e-05, + "loss": 0.0423, + "step": 44240 + }, + { + "epoch": 1.3658250509290697, + "grad_norm": 0.3288467824459076, + "learning_rate": 1.7687141050159246e-05, + "loss": 0.0407, + "step": 44250 + }, + { + "epoch": 1.3661337119575283, + "grad_norm": 0.5243694186210632, + "learning_rate": 1.7666116036257375e-05, + "loss": 0.0414, + "step": 44260 + }, + { + "epoch": 1.3664423729859867, + "grad_norm": 0.24158233404159546, + "learning_rate": 1.764510084453569e-05, + "loss": 0.0363, + "step": 44270 + }, + { + "epoch": 1.3667510340144453, + "grad_norm": 0.23958101868629456, + "learning_rate": 1.76240954813781e-05, + "loss": 0.0376, + "step": 44280 + }, + { + "epoch": 1.367059695042904, + "grad_norm": 0.4585687518119812, + "learning_rate": 1.7603099953165476e-05, + "loss": 0.0365, + "step": 44290 + }, + { + "epoch": 1.3673683560713625, + "grad_norm": 0.36407092213630676, + "learning_rate": 1.7582114266275683e-05, + "loss": 0.033, + "step": 44300 + }, + { + "epoch": 1.367677017099821, + "grad_norm": 0.5866534113883972, + "learning_rate": 1.756113842708364e-05, + "loss": 0.0333, + "step": 44310 + }, + { + "epoch": 1.3679856781282795, + "grad_norm": 0.3110520541667938, + "learning_rate": 1.7540172441961245e-05, + "loss": 0.0407, + "step": 44320 + }, + { + "epoch": 1.3682943391567381, + "grad_norm": 0.30264952778816223, + "learning_rate": 1.7519216317277387e-05, + "loss": 0.0348, + "step": 44330 + }, + { + "epoch": 1.3686030001851965, + "grad_norm": 0.30774542689323425, + "learning_rate": 1.7498270059398046e-05, + "loss": 0.0358, + "step": 44340 + }, + { + "epoch": 1.3689116612136552, + "grad_norm": 0.28552180528640747, + "learning_rate": 1.7477333674686062e-05, + "loss": 0.0369, + "step": 44350 + }, + { + "epoch": 1.3692203222421138, + "grad_norm": 0.2709135115146637, + "learning_rate": 1.745640716950142e-05, + "loss": 0.0369, + "step": 44360 + }, + { + "epoch": 1.3695289832705724, + "grad_norm": 0.32984650135040283, + "learning_rate": 1.7435490550201017e-05, + "loss": 0.0381, + "step": 44370 + }, + { + "epoch": 1.3698376442990308, + "grad_norm": 0.4497741162776947, + "learning_rate": 1.7414583823138762e-05, + "loss": 0.0346, + "step": 44380 + }, + { + "epoch": 1.3701463053274894, + "grad_norm": 0.48314785957336426, + "learning_rate": 1.739368699466558e-05, + "loss": 0.0346, + "step": 44390 + }, + { + "epoch": 1.3704549663559478, + "grad_norm": 0.300217866897583, + "learning_rate": 1.737280007112935e-05, + "loss": 0.0358, + "step": 44400 + }, + { + "epoch": 1.3707636273844064, + "grad_norm": 0.42810848355293274, + "learning_rate": 1.735192305887502e-05, + "loss": 0.0357, + "step": 44410 + }, + { + "epoch": 1.371072288412865, + "grad_norm": 0.3506160378456116, + "learning_rate": 1.733105596424441e-05, + "loss": 0.04, + "step": 44420 + }, + { + "epoch": 1.3713809494413236, + "grad_norm": 0.3341675400733948, + "learning_rate": 1.7310198793576437e-05, + "loss": 0.0372, + "step": 44430 + }, + { + "epoch": 1.3716896104697822, + "grad_norm": 0.2707105278968811, + "learning_rate": 1.7289351553206952e-05, + "loss": 0.0355, + "step": 44440 + }, + { + "epoch": 1.3719982714982406, + "grad_norm": 0.3823697865009308, + "learning_rate": 1.7268514249468788e-05, + "loss": 0.0374, + "step": 44450 + }, + { + "epoch": 1.3723069325266992, + "grad_norm": 0.31577757000923157, + "learning_rate": 1.7247686888691765e-05, + "loss": 0.0346, + "step": 44460 + }, + { + "epoch": 1.3726155935551576, + "grad_norm": 0.4459421932697296, + "learning_rate": 1.7226869477202694e-05, + "loss": 0.0383, + "step": 44470 + }, + { + "epoch": 1.3729242545836162, + "grad_norm": 0.2614741623401642, + "learning_rate": 1.7206062021325336e-05, + "loss": 0.0343, + "step": 44480 + }, + { + "epoch": 1.3732329156120748, + "grad_norm": 0.2660423815250397, + "learning_rate": 1.7185264527380502e-05, + "loss": 0.0325, + "step": 44490 + }, + { + "epoch": 1.3735415766405334, + "grad_norm": 0.414861798286438, + "learning_rate": 1.716447700168584e-05, + "loss": 0.0407, + "step": 44500 + }, + { + "epoch": 1.3738502376689918, + "grad_norm": 0.32675427198410034, + "learning_rate": 1.714369945055611e-05, + "loss": 0.0369, + "step": 44510 + }, + { + "epoch": 1.3741588986974504, + "grad_norm": 0.38168075680732727, + "learning_rate": 1.7122931880302968e-05, + "loss": 0.0335, + "step": 44520 + }, + { + "epoch": 1.374467559725909, + "grad_norm": 0.3398163318634033, + "learning_rate": 1.710217429723505e-05, + "loss": 0.038, + "step": 44530 + }, + { + "epoch": 1.3747762207543675, + "grad_norm": 0.3838254511356354, + "learning_rate": 1.7081426707657972e-05, + "loss": 0.0365, + "step": 44540 + }, + { + "epoch": 1.375084881782826, + "grad_norm": 0.36050307750701904, + "learning_rate": 1.7060689117874275e-05, + "loss": 0.0376, + "step": 44550 + }, + { + "epoch": 1.3753935428112847, + "grad_norm": 0.2853434979915619, + "learning_rate": 1.703996153418354e-05, + "loss": 0.038, + "step": 44560 + }, + { + "epoch": 1.3757022038397433, + "grad_norm": 0.4320106506347656, + "learning_rate": 1.7019243962882205e-05, + "loss": 0.0411, + "step": 44570 + }, + { + "epoch": 1.3760108648682017, + "grad_norm": 0.2672290802001953, + "learning_rate": 1.6998536410263754e-05, + "loss": 0.0359, + "step": 44580 + }, + { + "epoch": 1.3763195258966603, + "grad_norm": 0.3664899170398712, + "learning_rate": 1.6977838882618596e-05, + "loss": 0.0334, + "step": 44590 + }, + { + "epoch": 1.376628186925119, + "grad_norm": 0.30705103278160095, + "learning_rate": 1.6957151386234088e-05, + "loss": 0.0325, + "step": 44600 + }, + { + "epoch": 1.3769368479535773, + "grad_norm": 0.3529461920261383, + "learning_rate": 1.6936473927394536e-05, + "loss": 0.0348, + "step": 44610 + }, + { + "epoch": 1.377245508982036, + "grad_norm": 0.3987704813480377, + "learning_rate": 1.6915806512381222e-05, + "loss": 0.0337, + "step": 44620 + }, + { + "epoch": 1.3775541700104945, + "grad_norm": 0.3539005219936371, + "learning_rate": 1.6895149147472344e-05, + "loss": 0.0378, + "step": 44630 + }, + { + "epoch": 1.3778628310389531, + "grad_norm": 0.30323395133018494, + "learning_rate": 1.6874501838943073e-05, + "loss": 0.0331, + "step": 44640 + }, + { + "epoch": 1.3781714920674115, + "grad_norm": 0.3841431438922882, + "learning_rate": 1.6853864593065506e-05, + "loss": 0.0367, + "step": 44650 + }, + { + "epoch": 1.3784801530958701, + "grad_norm": 0.3304687440395355, + "learning_rate": 1.683323741610871e-05, + "loss": 0.0329, + "step": 44660 + }, + { + "epoch": 1.3787888141243287, + "grad_norm": 0.4022580683231354, + "learning_rate": 1.6812620314338674e-05, + "loss": 0.0349, + "step": 44670 + }, + { + "epoch": 1.3790974751527871, + "grad_norm": 0.263285756111145, + "learning_rate": 1.6792013294018326e-05, + "loss": 0.0428, + "step": 44680 + }, + { + "epoch": 1.3794061361812457, + "grad_norm": 0.38872018456459045, + "learning_rate": 1.6771416361407526e-05, + "loss": 0.0406, + "step": 44690 + }, + { + "epoch": 1.3797147972097044, + "grad_norm": 0.28740251064300537, + "learning_rate": 1.675082952276308e-05, + "loss": 0.0399, + "step": 44700 + }, + { + "epoch": 1.380023458238163, + "grad_norm": 0.4112080931663513, + "learning_rate": 1.6730252784338757e-05, + "loss": 0.0353, + "step": 44710 + }, + { + "epoch": 1.3803321192666214, + "grad_norm": 0.3080642521381378, + "learning_rate": 1.6709686152385166e-05, + "loss": 0.0416, + "step": 44720 + }, + { + "epoch": 1.38064078029508, + "grad_norm": 0.4031570851802826, + "learning_rate": 1.668912963314998e-05, + "loss": 0.037, + "step": 44730 + }, + { + "epoch": 1.3809494413235384, + "grad_norm": 0.3304741382598877, + "learning_rate": 1.6668583232877653e-05, + "loss": 0.037, + "step": 44740 + }, + { + "epoch": 1.381258102351997, + "grad_norm": 0.36057350039482117, + "learning_rate": 1.6648046957809698e-05, + "loss": 0.035, + "step": 44750 + }, + { + "epoch": 1.3815667633804556, + "grad_norm": 0.30028894543647766, + "learning_rate": 1.6627520814184462e-05, + "loss": 0.0347, + "step": 44760 + }, + { + "epoch": 1.3818754244089142, + "grad_norm": 0.27916160225868225, + "learning_rate": 1.660700480823726e-05, + "loss": 0.0357, + "step": 44770 + }, + { + "epoch": 1.3821840854373728, + "grad_norm": 0.6414340138435364, + "learning_rate": 1.65864989462003e-05, + "loss": 0.034, + "step": 44780 + }, + { + "epoch": 1.3824927464658312, + "grad_norm": 0.37771207094192505, + "learning_rate": 1.656600323430273e-05, + "loss": 0.0357, + "step": 44790 + }, + { + "epoch": 1.3828014074942898, + "grad_norm": 0.45046326518058777, + "learning_rate": 1.654551767877059e-05, + "loss": 0.0343, + "step": 44800 + }, + { + "epoch": 1.3831100685227482, + "grad_norm": 0.32098686695098877, + "learning_rate": 1.6525042285826874e-05, + "loss": 0.0411, + "step": 44810 + }, + { + "epoch": 1.3834187295512068, + "grad_norm": 0.2738197445869446, + "learning_rate": 1.6504577061691468e-05, + "loss": 0.0345, + "step": 44820 + }, + { + "epoch": 1.3837273905796654, + "grad_norm": 0.4151959717273712, + "learning_rate": 1.6484122012581143e-05, + "loss": 0.0352, + "step": 44830 + }, + { + "epoch": 1.384036051608124, + "grad_norm": 0.3174365162849426, + "learning_rate": 1.6463677144709623e-05, + "loss": 0.0391, + "step": 44840 + }, + { + "epoch": 1.3843447126365824, + "grad_norm": 0.4029141664505005, + "learning_rate": 1.6443242464287493e-05, + "loss": 0.0346, + "step": 44850 + }, + { + "epoch": 1.384653373665041, + "grad_norm": 0.3532199263572693, + "learning_rate": 1.642281797752232e-05, + "loss": 0.0355, + "step": 44860 + }, + { + "epoch": 1.3849620346934997, + "grad_norm": 0.34262773394584656, + "learning_rate": 1.6402403690618456e-05, + "loss": 0.0325, + "step": 44870 + }, + { + "epoch": 1.385270695721958, + "grad_norm": 0.3030531108379364, + "learning_rate": 1.6381999609777295e-05, + "loss": 0.0364, + "step": 44880 + }, + { + "epoch": 1.3855793567504167, + "grad_norm": 0.4829317033290863, + "learning_rate": 1.6361605741196983e-05, + "loss": 0.0403, + "step": 44890 + }, + { + "epoch": 1.3858880177788753, + "grad_norm": 0.3170856535434723, + "learning_rate": 1.63412220910727e-05, + "loss": 0.0389, + "step": 44900 + }, + { + "epoch": 1.3861966788073339, + "grad_norm": 0.4072454571723938, + "learning_rate": 1.6320848665596433e-05, + "loss": 0.0384, + "step": 44910 + }, + { + "epoch": 1.3865053398357923, + "grad_norm": 0.33906444907188416, + "learning_rate": 1.6300485470957095e-05, + "loss": 0.0381, + "step": 44920 + }, + { + "epoch": 1.3868140008642509, + "grad_norm": 0.28033286333084106, + "learning_rate": 1.6280132513340483e-05, + "loss": 0.0382, + "step": 44930 + }, + { + "epoch": 1.3871226618927095, + "grad_norm": 0.38639387488365173, + "learning_rate": 1.62597897989293e-05, + "loss": 0.0368, + "step": 44940 + }, + { + "epoch": 1.3874313229211679, + "grad_norm": 0.27066880464553833, + "learning_rate": 1.623945733390309e-05, + "loss": 0.0341, + "step": 44950 + }, + { + "epoch": 1.3877399839496265, + "grad_norm": 0.23810312151908875, + "learning_rate": 1.6219135124438374e-05, + "loss": 0.0323, + "step": 44960 + }, + { + "epoch": 1.3880486449780851, + "grad_norm": 0.2910325229167938, + "learning_rate": 1.6198823176708465e-05, + "loss": 0.0313, + "step": 44970 + }, + { + "epoch": 1.3883573060065437, + "grad_norm": 0.3225690722465515, + "learning_rate": 1.6178521496883613e-05, + "loss": 0.0329, + "step": 44980 + }, + { + "epoch": 1.3886659670350021, + "grad_norm": 0.25227510929107666, + "learning_rate": 1.6158230091130926e-05, + "loss": 0.0346, + "step": 44990 + }, + { + "epoch": 1.3889746280634607, + "grad_norm": 0.3328612148761749, + "learning_rate": 1.613794896561438e-05, + "loss": 0.0379, + "step": 45000 + }, + { + "epoch": 1.3892832890919191, + "grad_norm": 0.25952717661857605, + "learning_rate": 1.6117678126494894e-05, + "loss": 0.0399, + "step": 45010 + }, + { + "epoch": 1.3895919501203777, + "grad_norm": 0.3422434628009796, + "learning_rate": 1.6097417579930153e-05, + "loss": 0.0384, + "step": 45020 + }, + { + "epoch": 1.3899006111488363, + "grad_norm": 0.2943617105484009, + "learning_rate": 1.6077167332074834e-05, + "loss": 0.0349, + "step": 45030 + }, + { + "epoch": 1.390209272177295, + "grad_norm": 0.3786255419254303, + "learning_rate": 1.605692738908037e-05, + "loss": 0.039, + "step": 45040 + }, + { + "epoch": 1.3905179332057536, + "grad_norm": 0.34198176860809326, + "learning_rate": 1.6036697757095176e-05, + "loss": 0.0364, + "step": 45050 + }, + { + "epoch": 1.390826594234212, + "grad_norm": 0.25723323225975037, + "learning_rate": 1.6016478442264428e-05, + "loss": 0.0352, + "step": 45060 + }, + { + "epoch": 1.3911352552626706, + "grad_norm": 0.35814908146858215, + "learning_rate": 1.599626945073026e-05, + "loss": 0.04, + "step": 45070 + }, + { + "epoch": 1.391443916291129, + "grad_norm": 0.4661754369735718, + "learning_rate": 1.597607078863162e-05, + "loss": 0.0347, + "step": 45080 + }, + { + "epoch": 1.3917525773195876, + "grad_norm": 0.2784949243068695, + "learning_rate": 1.595588246210432e-05, + "loss": 0.0381, + "step": 45090 + }, + { + "epoch": 1.3920612383480462, + "grad_norm": 0.26293495297431946, + "learning_rate": 1.5935704477281048e-05, + "loss": 0.0357, + "step": 45100 + }, + { + "epoch": 1.3923698993765048, + "grad_norm": 0.3468224108219147, + "learning_rate": 1.5915536840291323e-05, + "loss": 0.0406, + "step": 45110 + }, + { + "epoch": 1.3926785604049634, + "grad_norm": 0.2373930960893631, + "learning_rate": 1.5895379557261576e-05, + "loss": 0.0348, + "step": 45120 + }, + { + "epoch": 1.3929872214334218, + "grad_norm": 0.33779919147491455, + "learning_rate": 1.5875232634315033e-05, + "loss": 0.0381, + "step": 45130 + }, + { + "epoch": 1.3932958824618804, + "grad_norm": 0.748822033405304, + "learning_rate": 1.5855096077571812e-05, + "loss": 0.0391, + "step": 45140 + }, + { + "epoch": 1.3936045434903388, + "grad_norm": 0.37610551714897156, + "learning_rate": 1.5834969893148855e-05, + "loss": 0.0344, + "step": 45150 + }, + { + "epoch": 1.3939132045187974, + "grad_norm": 0.34061670303344727, + "learning_rate": 1.581485408715997e-05, + "loss": 0.0349, + "step": 45160 + }, + { + "epoch": 1.394221865547256, + "grad_norm": 0.3414478600025177, + "learning_rate": 1.5794748665715785e-05, + "loss": 0.0334, + "step": 45170 + }, + { + "epoch": 1.3945305265757146, + "grad_norm": 0.26768603920936584, + "learning_rate": 1.5774653634923857e-05, + "loss": 0.0324, + "step": 45180 + }, + { + "epoch": 1.394839187604173, + "grad_norm": 0.2570499777793884, + "learning_rate": 1.575456900088845e-05, + "loss": 0.0378, + "step": 45190 + }, + { + "epoch": 1.3951478486326316, + "grad_norm": 0.31677594780921936, + "learning_rate": 1.5734494769710816e-05, + "loss": 0.0363, + "step": 45200 + }, + { + "epoch": 1.3954565096610902, + "grad_norm": 0.3567831218242645, + "learning_rate": 1.5714430947488912e-05, + "loss": 0.0339, + "step": 45210 + }, + { + "epoch": 1.3957651706895486, + "grad_norm": 0.2945173680782318, + "learning_rate": 1.5694377540317645e-05, + "loss": 0.0386, + "step": 45220 + }, + { + "epoch": 1.3960738317180073, + "grad_norm": 0.33976590633392334, + "learning_rate": 1.5674334554288694e-05, + "loss": 0.0331, + "step": 45230 + }, + { + "epoch": 1.3963824927464659, + "grad_norm": 0.3598896265029907, + "learning_rate": 1.5654301995490582e-05, + "loss": 0.0346, + "step": 45240 + }, + { + "epoch": 1.3966911537749245, + "grad_norm": 0.3259637951850891, + "learning_rate": 1.5634279870008685e-05, + "loss": 0.0368, + "step": 45250 + }, + { + "epoch": 1.3969998148033829, + "grad_norm": 0.3106662929058075, + "learning_rate": 1.5614268183925174e-05, + "loss": 0.039, + "step": 45260 + }, + { + "epoch": 1.3973084758318415, + "grad_norm": 0.37372103333473206, + "learning_rate": 1.5594266943319097e-05, + "loss": 0.0382, + "step": 45270 + }, + { + "epoch": 1.3976171368603, + "grad_norm": 0.3142867684364319, + "learning_rate": 1.5574276154266294e-05, + "loss": 0.0357, + "step": 45280 + }, + { + "epoch": 1.3979257978887585, + "grad_norm": 0.28863632678985596, + "learning_rate": 1.5554295822839437e-05, + "loss": 0.0389, + "step": 45290 + }, + { + "epoch": 1.398234458917217, + "grad_norm": 0.5321882367134094, + "learning_rate": 1.5534325955108025e-05, + "loss": 0.0343, + "step": 45300 + }, + { + "epoch": 1.3985431199456757, + "grad_norm": 0.32839423418045044, + "learning_rate": 1.5514366557138373e-05, + "loss": 0.0395, + "step": 45310 + }, + { + "epoch": 1.3988517809741343, + "grad_norm": 0.3568706214427948, + "learning_rate": 1.5494417634993602e-05, + "loss": 0.0341, + "step": 45320 + }, + { + "epoch": 1.3991604420025927, + "grad_norm": 0.30906495451927185, + "learning_rate": 1.547447919473372e-05, + "loss": 0.0342, + "step": 45330 + }, + { + "epoch": 1.3994691030310513, + "grad_norm": 0.28814566135406494, + "learning_rate": 1.5454551242415434e-05, + "loss": 0.0378, + "step": 45340 + }, + { + "epoch": 1.3997777640595097, + "grad_norm": 0.4723045229911804, + "learning_rate": 1.543463378409239e-05, + "loss": 0.0353, + "step": 45350 + }, + { + "epoch": 1.4000864250879683, + "grad_norm": 0.37055569887161255, + "learning_rate": 1.541472682581493e-05, + "loss": 0.032, + "step": 45360 + }, + { + "epoch": 1.400395086116427, + "grad_norm": 0.28390049934387207, + "learning_rate": 1.5394830373630298e-05, + "loss": 0.0395, + "step": 45370 + }, + { + "epoch": 1.4007037471448855, + "grad_norm": 0.41694411635398865, + "learning_rate": 1.5374944433582506e-05, + "loss": 0.0362, + "step": 45380 + }, + { + "epoch": 1.4010124081733442, + "grad_norm": 0.31150364875793457, + "learning_rate": 1.5355069011712375e-05, + "loss": 0.0357, + "step": 45390 + }, + { + "epoch": 1.4013210692018025, + "grad_norm": 0.32944026589393616, + "learning_rate": 1.5335204114057526e-05, + "loss": 0.0339, + "step": 45400 + }, + { + "epoch": 1.4016297302302612, + "grad_norm": 0.39902764558792114, + "learning_rate": 1.5315349746652387e-05, + "loss": 0.0348, + "step": 45410 + }, + { + "epoch": 1.4019383912587196, + "grad_norm": 0.3013714551925659, + "learning_rate": 1.5295505915528212e-05, + "loss": 0.0368, + "step": 45420 + }, + { + "epoch": 1.4022470522871782, + "grad_norm": 0.3622114658355713, + "learning_rate": 1.5275672626713024e-05, + "loss": 0.0371, + "step": 45430 + }, + { + "epoch": 1.4025557133156368, + "grad_norm": 0.463293194770813, + "learning_rate": 1.5255849886231643e-05, + "loss": 0.0334, + "step": 45440 + }, + { + "epoch": 1.4028643743440954, + "grad_norm": 0.3960438072681427, + "learning_rate": 1.523603770010571e-05, + "loss": 0.0378, + "step": 45450 + }, + { + "epoch": 1.4031730353725538, + "grad_norm": 0.2797049880027771, + "learning_rate": 1.521623607435363e-05, + "loss": 0.0409, + "step": 45460 + }, + { + "epoch": 1.4034816964010124, + "grad_norm": 0.29399505257606506, + "learning_rate": 1.5196445014990612e-05, + "loss": 0.0362, + "step": 45470 + }, + { + "epoch": 1.403790357429471, + "grad_norm": 0.4046132564544678, + "learning_rate": 1.5176664528028672e-05, + "loss": 0.0371, + "step": 45480 + }, + { + "epoch": 1.4040990184579294, + "grad_norm": 0.3525623381137848, + "learning_rate": 1.5156894619476574e-05, + "loss": 0.0351, + "step": 45490 + }, + { + "epoch": 1.404407679486388, + "grad_norm": 0.3658609986305237, + "learning_rate": 1.5137135295339938e-05, + "loss": 0.0329, + "step": 45500 + }, + { + "epoch": 1.4047163405148466, + "grad_norm": 0.3340725004673004, + "learning_rate": 1.5117386561621073e-05, + "loss": 0.0369, + "step": 45510 + }, + { + "epoch": 1.4050250015433052, + "grad_norm": 0.39254021644592285, + "learning_rate": 1.5097648424319167e-05, + "loss": 0.0399, + "step": 45520 + }, + { + "epoch": 1.4053336625717636, + "grad_norm": 0.3695322275161743, + "learning_rate": 1.5077920889430119e-05, + "loss": 0.0347, + "step": 45530 + }, + { + "epoch": 1.4056423236002222, + "grad_norm": 0.2775259017944336, + "learning_rate": 1.5058203962946644e-05, + "loss": 0.0352, + "step": 45540 + }, + { + "epoch": 1.4059509846286808, + "grad_norm": 0.27830880880355835, + "learning_rate": 1.503849765085822e-05, + "loss": 0.0379, + "step": 45550 + }, + { + "epoch": 1.4062596456571392, + "grad_norm": 0.2506062388420105, + "learning_rate": 1.501880195915109e-05, + "loss": 0.0353, + "step": 45560 + }, + { + "epoch": 1.4065683066855978, + "grad_norm": 0.3221781849861145, + "learning_rate": 1.499911689380833e-05, + "loss": 0.0415, + "step": 45570 + }, + { + "epoch": 1.4068769677140565, + "grad_norm": 0.3646438717842102, + "learning_rate": 1.4979442460809683e-05, + "loss": 0.0329, + "step": 45580 + }, + { + "epoch": 1.407185628742515, + "grad_norm": 0.30325451493263245, + "learning_rate": 1.4959778666131763e-05, + "loss": 0.0374, + "step": 45590 + }, + { + "epoch": 1.4074942897709735, + "grad_norm": 0.33118414878845215, + "learning_rate": 1.4940125515747905e-05, + "loss": 0.0377, + "step": 45600 + }, + { + "epoch": 1.407802950799432, + "grad_norm": 0.27071908116340637, + "learning_rate": 1.4920483015628211e-05, + "loss": 0.0333, + "step": 45610 + }, + { + "epoch": 1.4081116118278907, + "grad_norm": 0.30821147561073303, + "learning_rate": 1.490085117173956e-05, + "loss": 0.036, + "step": 45620 + }, + { + "epoch": 1.408420272856349, + "grad_norm": 0.28878000378608704, + "learning_rate": 1.488122999004558e-05, + "loss": 0.0353, + "step": 45630 + }, + { + "epoch": 1.4087289338848077, + "grad_norm": 0.36030471324920654, + "learning_rate": 1.486161947650666e-05, + "loss": 0.0375, + "step": 45640 + }, + { + "epoch": 1.4090375949132663, + "grad_norm": 0.3872099220752716, + "learning_rate": 1.4842019637079995e-05, + "loss": 0.0368, + "step": 45650 + }, + { + "epoch": 1.409346255941725, + "grad_norm": 0.2792753279209137, + "learning_rate": 1.482243047771944e-05, + "loss": 0.037, + "step": 45660 + }, + { + "epoch": 1.4096549169701833, + "grad_norm": 0.331073135137558, + "learning_rate": 1.4802852004375712e-05, + "loss": 0.0336, + "step": 45670 + }, + { + "epoch": 1.409963577998642, + "grad_norm": 0.35529929399490356, + "learning_rate": 1.4783284222996218e-05, + "loss": 0.0359, + "step": 45680 + }, + { + "epoch": 1.4102722390271003, + "grad_norm": 0.3125388026237488, + "learning_rate": 1.4763727139525135e-05, + "loss": 0.0344, + "step": 45690 + }, + { + "epoch": 1.410580900055559, + "grad_norm": 0.5808215737342834, + "learning_rate": 1.4744180759903392e-05, + "loss": 0.0331, + "step": 45700 + }, + { + "epoch": 1.4108895610840175, + "grad_norm": 0.48037609457969666, + "learning_rate": 1.4724645090068635e-05, + "loss": 0.0413, + "step": 45710 + }, + { + "epoch": 1.4111982221124761, + "grad_norm": 0.2535495460033417, + "learning_rate": 1.4705120135955341e-05, + "loss": 0.0311, + "step": 45720 + }, + { + "epoch": 1.4115068831409348, + "grad_norm": 0.30838799476623535, + "learning_rate": 1.4685605903494614e-05, + "loss": 0.0338, + "step": 45730 + }, + { + "epoch": 1.4118155441693931, + "grad_norm": 2.0926530361175537, + "learning_rate": 1.46661023986144e-05, + "loss": 0.0312, + "step": 45740 + }, + { + "epoch": 1.4121242051978518, + "grad_norm": 0.3481905460357666, + "learning_rate": 1.4646609627239344e-05, + "loss": 0.0342, + "step": 45750 + }, + { + "epoch": 1.4124328662263101, + "grad_norm": 0.7341243624687195, + "learning_rate": 1.4627127595290835e-05, + "loss": 0.0366, + "step": 45760 + }, + { + "epoch": 1.4127415272547688, + "grad_norm": 0.26489779353141785, + "learning_rate": 1.460765630868699e-05, + "loss": 0.0334, + "step": 45770 + }, + { + "epoch": 1.4130501882832274, + "grad_norm": 0.4583755433559418, + "learning_rate": 1.4588195773342678e-05, + "loss": 0.0381, + "step": 45780 + }, + { + "epoch": 1.413358849311686, + "grad_norm": 0.3012583553791046, + "learning_rate": 1.4568745995169485e-05, + "loss": 0.0355, + "step": 45790 + }, + { + "epoch": 1.4136675103401444, + "grad_norm": 0.3626764714717865, + "learning_rate": 1.4549306980075778e-05, + "loss": 0.0347, + "step": 45800 + }, + { + "epoch": 1.413976171368603, + "grad_norm": 0.3458002507686615, + "learning_rate": 1.4529878733966557e-05, + "loss": 0.0375, + "step": 45810 + }, + { + "epoch": 1.4142848323970616, + "grad_norm": 0.3269500732421875, + "learning_rate": 1.4510461262743658e-05, + "loss": 0.0351, + "step": 45820 + }, + { + "epoch": 1.41459349342552, + "grad_norm": 0.25330907106399536, + "learning_rate": 1.4491054572305585e-05, + "loss": 0.0352, + "step": 45830 + }, + { + "epoch": 1.4149021544539786, + "grad_norm": 0.3923347294330597, + "learning_rate": 1.4471658668547566e-05, + "loss": 0.035, + "step": 45840 + }, + { + "epoch": 1.4152108154824372, + "grad_norm": 0.4158909320831299, + "learning_rate": 1.4452273557361579e-05, + "loss": 0.035, + "step": 45850 + }, + { + "epoch": 1.4155194765108958, + "grad_norm": 0.30342477560043335, + "learning_rate": 1.4432899244636282e-05, + "loss": 0.0312, + "step": 45860 + }, + { + "epoch": 1.4158281375393542, + "grad_norm": 0.4847378432750702, + "learning_rate": 1.4413535736257134e-05, + "loss": 0.0364, + "step": 45870 + }, + { + "epoch": 1.4161367985678128, + "grad_norm": 0.410035103559494, + "learning_rate": 1.439418303810619e-05, + "loss": 0.0373, + "step": 45880 + }, + { + "epoch": 1.4164454595962714, + "grad_norm": 0.4507884383201599, + "learning_rate": 1.4374841156062352e-05, + "loss": 0.0337, + "step": 45890 + }, + { + "epoch": 1.4167541206247298, + "grad_norm": 0.34790515899658203, + "learning_rate": 1.4355510096001112e-05, + "loss": 0.04, + "step": 45900 + }, + { + "epoch": 1.4170627816531884, + "grad_norm": 0.3615141212940216, + "learning_rate": 1.4336189863794786e-05, + "loss": 0.0377, + "step": 45910 + }, + { + "epoch": 1.417371442681647, + "grad_norm": 0.2726093828678131, + "learning_rate": 1.4316880465312327e-05, + "loss": 0.0364, + "step": 45920 + }, + { + "epoch": 1.4176801037101057, + "grad_norm": 0.3150503635406494, + "learning_rate": 1.4297581906419426e-05, + "loss": 0.0338, + "step": 45930 + }, + { + "epoch": 1.417988764738564, + "grad_norm": 0.340261310338974, + "learning_rate": 1.4278294192978475e-05, + "loss": 0.0366, + "step": 45940 + }, + { + "epoch": 1.4182974257670227, + "grad_norm": 0.3362569510936737, + "learning_rate": 1.4259017330848574e-05, + "loss": 0.0335, + "step": 45950 + }, + { + "epoch": 1.4186060867954813, + "grad_norm": 0.4099630117416382, + "learning_rate": 1.4239751325885498e-05, + "loss": 0.0345, + "step": 45960 + }, + { + "epoch": 1.4189147478239397, + "grad_norm": 0.2725518047809601, + "learning_rate": 1.4220496183941795e-05, + "loss": 0.0354, + "step": 45970 + }, + { + "epoch": 1.4192234088523983, + "grad_norm": 0.6472665667533875, + "learning_rate": 1.4201251910866648e-05, + "loss": 0.0419, + "step": 45980 + }, + { + "epoch": 1.419532069880857, + "grad_norm": 0.3261873126029968, + "learning_rate": 1.4182018512505957e-05, + "loss": 0.0343, + "step": 45990 + }, + { + "epoch": 1.4198407309093155, + "grad_norm": 0.3658423125743866, + "learning_rate": 1.4162795994702327e-05, + "loss": 0.0393, + "step": 46000 + }, + { + "epoch": 1.420149391937774, + "grad_norm": 0.4528888761997223, + "learning_rate": 1.4143584363295032e-05, + "loss": 0.0332, + "step": 46010 + }, + { + "epoch": 1.4204580529662325, + "grad_norm": 0.26646554470062256, + "learning_rate": 1.4124383624120101e-05, + "loss": 0.0381, + "step": 46020 + }, + { + "epoch": 1.420766713994691, + "grad_norm": 0.4518396258354187, + "learning_rate": 1.4105193783010151e-05, + "loss": 0.0359, + "step": 46030 + }, + { + "epoch": 1.4210753750231495, + "grad_norm": 0.3168582618236542, + "learning_rate": 1.4086014845794621e-05, + "loss": 0.0396, + "step": 46040 + }, + { + "epoch": 1.4213840360516081, + "grad_norm": 0.38549619913101196, + "learning_rate": 1.4066846818299489e-05, + "loss": 0.0365, + "step": 46050 + }, + { + "epoch": 1.4216926970800667, + "grad_norm": 0.29966720938682556, + "learning_rate": 1.4047689706347555e-05, + "loss": 0.035, + "step": 46060 + }, + { + "epoch": 1.4220013581085253, + "grad_norm": 0.33128342032432556, + "learning_rate": 1.402854351575822e-05, + "loss": 0.0348, + "step": 46070 + }, + { + "epoch": 1.4223100191369837, + "grad_norm": 0.3185531497001648, + "learning_rate": 1.4009408252347588e-05, + "loss": 0.0341, + "step": 46080 + }, + { + "epoch": 1.4226186801654424, + "grad_norm": 0.3409533202648163, + "learning_rate": 1.399028392192846e-05, + "loss": 0.0321, + "step": 46090 + }, + { + "epoch": 1.4229273411939007, + "grad_norm": 0.2878316044807434, + "learning_rate": 1.397117053031029e-05, + "loss": 0.0331, + "step": 46100 + }, + { + "epoch": 1.4232360022223594, + "grad_norm": 0.3562215268611908, + "learning_rate": 1.3952068083299213e-05, + "loss": 0.038, + "step": 46110 + }, + { + "epoch": 1.423544663250818, + "grad_norm": 0.3091203570365906, + "learning_rate": 1.3932976586698082e-05, + "loss": 0.0366, + "step": 46120 + }, + { + "epoch": 1.4238533242792766, + "grad_norm": 0.3689088225364685, + "learning_rate": 1.3913896046306363e-05, + "loss": 0.0346, + "step": 46130 + }, + { + "epoch": 1.424161985307735, + "grad_norm": 0.32321786880493164, + "learning_rate": 1.389482646792023e-05, + "loss": 0.0347, + "step": 46140 + }, + { + "epoch": 1.4244706463361936, + "grad_norm": 0.27991166710853577, + "learning_rate": 1.387576785733251e-05, + "loss": 0.0357, + "step": 46150 + }, + { + "epoch": 1.4247793073646522, + "grad_norm": 0.32499605417251587, + "learning_rate": 1.3856720220332703e-05, + "loss": 0.0354, + "step": 46160 + }, + { + "epoch": 1.4250879683931106, + "grad_norm": 0.27070656418800354, + "learning_rate": 1.383768356270701e-05, + "loss": 0.0355, + "step": 46170 + }, + { + "epoch": 1.4253966294215692, + "grad_norm": 0.3219636380672455, + "learning_rate": 1.3818657890238207e-05, + "loss": 0.0394, + "step": 46180 + }, + { + "epoch": 1.4257052904500278, + "grad_norm": 0.23914146423339844, + "learning_rate": 1.3799643208705859e-05, + "loss": 0.0313, + "step": 46190 + }, + { + "epoch": 1.4260139514784864, + "grad_norm": 0.6966075301170349, + "learning_rate": 1.3780639523886058e-05, + "loss": 0.037, + "step": 46200 + }, + { + "epoch": 1.4263226125069448, + "grad_norm": 1.3668465614318848, + "learning_rate": 1.3761646841551668e-05, + "loss": 0.0362, + "step": 46210 + }, + { + "epoch": 1.4266312735354034, + "grad_norm": 0.29698753356933594, + "learning_rate": 1.3742665167472146e-05, + "loss": 0.033, + "step": 46220 + }, + { + "epoch": 1.426939934563862, + "grad_norm": 0.33460187911987305, + "learning_rate": 1.372369450741363e-05, + "loss": 0.0355, + "step": 46230 + }, + { + "epoch": 1.4272485955923204, + "grad_norm": 0.28311052918434143, + "learning_rate": 1.3704734867138901e-05, + "loss": 0.0374, + "step": 46240 + }, + { + "epoch": 1.427557256620779, + "grad_norm": 0.24574187397956848, + "learning_rate": 1.36857862524074e-05, + "loss": 0.0324, + "step": 46250 + }, + { + "epoch": 1.4278659176492376, + "grad_norm": 0.4272107481956482, + "learning_rate": 1.3666848668975213e-05, + "loss": 0.0348, + "step": 46260 + }, + { + "epoch": 1.4281745786776963, + "grad_norm": 0.3573107421398163, + "learning_rate": 1.3647922122595063e-05, + "loss": 0.0346, + "step": 46270 + }, + { + "epoch": 1.4284832397061547, + "grad_norm": 0.21981047093868256, + "learning_rate": 1.3629006619016366e-05, + "loss": 0.0342, + "step": 46280 + }, + { + "epoch": 1.4287919007346133, + "grad_norm": 0.22446025907993317, + "learning_rate": 1.3610102163985139e-05, + "loss": 0.0345, + "step": 46290 + }, + { + "epoch": 1.4291005617630719, + "grad_norm": 0.24895913898944855, + "learning_rate": 1.3591208763244057e-05, + "loss": 0.036, + "step": 46300 + }, + { + "epoch": 1.4294092227915303, + "grad_norm": 0.5000686049461365, + "learning_rate": 1.3572326422532428e-05, + "loss": 0.0324, + "step": 46310 + }, + { + "epoch": 1.4297178838199889, + "grad_norm": 0.35630184412002563, + "learning_rate": 1.355345514758622e-05, + "loss": 0.0387, + "step": 46320 + }, + { + "epoch": 1.4300265448484475, + "grad_norm": 0.44205060601234436, + "learning_rate": 1.3534594944138007e-05, + "loss": 0.0334, + "step": 46330 + }, + { + "epoch": 1.430335205876906, + "grad_norm": 0.4046424925327301, + "learning_rate": 1.3515745817917069e-05, + "loss": 0.0398, + "step": 46340 + }, + { + "epoch": 1.4306438669053645, + "grad_norm": 0.47785821557044983, + "learning_rate": 1.3496907774649208e-05, + "loss": 0.033, + "step": 46350 + }, + { + "epoch": 1.430952527933823, + "grad_norm": 0.2234039604663849, + "learning_rate": 1.3478080820056987e-05, + "loss": 0.0329, + "step": 46360 + }, + { + "epoch": 1.4312611889622815, + "grad_norm": 0.27502238750457764, + "learning_rate": 1.3459264959859474e-05, + "loss": 0.0368, + "step": 46370 + }, + { + "epoch": 1.43156984999074, + "grad_norm": 0.39739611744880676, + "learning_rate": 1.3440460199772487e-05, + "loss": 0.0354, + "step": 46380 + }, + { + "epoch": 1.4318785110191987, + "grad_norm": 0.2722068727016449, + "learning_rate": 1.3421666545508382e-05, + "loss": 0.0351, + "step": 46390 + }, + { + "epoch": 1.4321871720476573, + "grad_norm": 0.34242120385169983, + "learning_rate": 1.3402884002776194e-05, + "loss": 0.0386, + "step": 46400 + }, + { + "epoch": 1.432495833076116, + "grad_norm": 0.2518649399280548, + "learning_rate": 1.3384112577281555e-05, + "loss": 0.0381, + "step": 46410 + }, + { + "epoch": 1.4328044941045743, + "grad_norm": 0.645542323589325, + "learning_rate": 1.3365352274726711e-05, + "loss": 0.0336, + "step": 46420 + }, + { + "epoch": 1.433113155133033, + "grad_norm": 0.3736375868320465, + "learning_rate": 1.3346603100810578e-05, + "loss": 0.0371, + "step": 46430 + }, + { + "epoch": 1.4334218161614913, + "grad_norm": 0.520109236240387, + "learning_rate": 1.3327865061228645e-05, + "loss": 0.0359, + "step": 46440 + }, + { + "epoch": 1.43373047718995, + "grad_norm": 0.45480623841285706, + "learning_rate": 1.330913816167304e-05, + "loss": 0.036, + "step": 46450 + }, + { + "epoch": 1.4340391382184086, + "grad_norm": 0.24092836678028107, + "learning_rate": 1.3290422407832492e-05, + "loss": 0.0359, + "step": 46460 + }, + { + "epoch": 1.4343477992468672, + "grad_norm": 0.27796071767807007, + "learning_rate": 1.3271717805392354e-05, + "loss": 0.0396, + "step": 46470 + }, + { + "epoch": 1.4346564602753256, + "grad_norm": 0.3294979929924011, + "learning_rate": 1.3253024360034582e-05, + "loss": 0.0414, + "step": 46480 + }, + { + "epoch": 1.4349651213037842, + "grad_norm": 0.306343674659729, + "learning_rate": 1.323434207743779e-05, + "loss": 0.032, + "step": 46490 + }, + { + "epoch": 1.4352737823322428, + "grad_norm": 0.31411129236221313, + "learning_rate": 1.3215670963277105e-05, + "loss": 0.0355, + "step": 46500 + }, + { + "epoch": 1.4355824433607012, + "grad_norm": 0.252153605222702, + "learning_rate": 1.3197011023224376e-05, + "loss": 0.0353, + "step": 46510 + }, + { + "epoch": 1.4358911043891598, + "grad_norm": 0.28426671028137207, + "learning_rate": 1.3178362262947941e-05, + "loss": 0.0332, + "step": 46520 + }, + { + "epoch": 1.4361997654176184, + "grad_norm": 0.40520966053009033, + "learning_rate": 1.3159724688112845e-05, + "loss": 0.0323, + "step": 46530 + }, + { + "epoch": 1.436508426446077, + "grad_norm": 0.34950903058052063, + "learning_rate": 1.3141098304380683e-05, + "loss": 0.0402, + "step": 46540 + }, + { + "epoch": 1.4368170874745354, + "grad_norm": 0.3277672827243805, + "learning_rate": 1.3122483117409651e-05, + "loss": 0.0367, + "step": 46550 + }, + { + "epoch": 1.437125748502994, + "grad_norm": 0.3171308934688568, + "learning_rate": 1.3103879132854552e-05, + "loss": 0.0375, + "step": 46560 + }, + { + "epoch": 1.4374344095314526, + "grad_norm": 0.3657664656639099, + "learning_rate": 1.3085286356366771e-05, + "loss": 0.0363, + "step": 46570 + }, + { + "epoch": 1.437743070559911, + "grad_norm": 0.35095641016960144, + "learning_rate": 1.3066704793594337e-05, + "loss": 0.0339, + "step": 46580 + }, + { + "epoch": 1.4380517315883696, + "grad_norm": 0.24223539233207703, + "learning_rate": 1.3048134450181816e-05, + "loss": 0.0326, + "step": 46590 + }, + { + "epoch": 1.4383603926168282, + "grad_norm": 0.3158322274684906, + "learning_rate": 1.3029575331770394e-05, + "loss": 0.0378, + "step": 46600 + }, + { + "epoch": 1.4386690536452869, + "grad_norm": 0.26400625705718994, + "learning_rate": 1.3011027443997837e-05, + "loss": 0.0339, + "step": 46610 + }, + { + "epoch": 1.4389777146737452, + "grad_norm": 0.4154921770095825, + "learning_rate": 1.2992490792498507e-05, + "loss": 0.0364, + "step": 46620 + }, + { + "epoch": 1.4392863757022039, + "grad_norm": 0.23644186556339264, + "learning_rate": 1.297396538290333e-05, + "loss": 0.0336, + "step": 46630 + }, + { + "epoch": 1.4395950367306625, + "grad_norm": 0.40523114800453186, + "learning_rate": 1.2955451220839888e-05, + "loss": 0.0351, + "step": 46640 + }, + { + "epoch": 1.4399036977591209, + "grad_norm": 0.30462443828582764, + "learning_rate": 1.2936948311932223e-05, + "loss": 0.0361, + "step": 46650 + }, + { + "epoch": 1.4402123587875795, + "grad_norm": 0.329549640417099, + "learning_rate": 1.2918456661801104e-05, + "loss": 0.0357, + "step": 46660 + }, + { + "epoch": 1.440521019816038, + "grad_norm": 0.22736775875091553, + "learning_rate": 1.2899976276063736e-05, + "loss": 0.0338, + "step": 46670 + }, + { + "epoch": 1.4408296808444967, + "grad_norm": 0.34228062629699707, + "learning_rate": 1.2881507160334022e-05, + "loss": 0.0396, + "step": 46680 + }, + { + "epoch": 1.441138341872955, + "grad_norm": 0.3715246319770813, + "learning_rate": 1.286304932022238e-05, + "loss": 0.034, + "step": 46690 + }, + { + "epoch": 1.4414470029014137, + "grad_norm": 0.31200528144836426, + "learning_rate": 1.2844602761335806e-05, + "loss": 0.0328, + "step": 46700 + }, + { + "epoch": 1.441755663929872, + "grad_norm": 0.22929348051548004, + "learning_rate": 1.2826167489277885e-05, + "loss": 0.0313, + "step": 46710 + }, + { + "epoch": 1.4420643249583307, + "grad_norm": 0.2799290418624878, + "learning_rate": 1.2807743509648745e-05, + "loss": 0.0323, + "step": 46720 + }, + { + "epoch": 1.4423729859867893, + "grad_norm": 0.340018093585968, + "learning_rate": 1.2789330828045149e-05, + "loss": 0.0414, + "step": 46730 + }, + { + "epoch": 1.442681647015248, + "grad_norm": 0.31338992714881897, + "learning_rate": 1.2770929450060332e-05, + "loss": 0.0292, + "step": 46740 + }, + { + "epoch": 1.4429903080437065, + "grad_norm": 0.36590391397476196, + "learning_rate": 1.2752539381284184e-05, + "loss": 0.0377, + "step": 46750 + }, + { + "epoch": 1.443298969072165, + "grad_norm": 0.24196454882621765, + "learning_rate": 1.273416062730311e-05, + "loss": 0.0352, + "step": 46760 + }, + { + "epoch": 1.4436076301006235, + "grad_norm": 0.25930535793304443, + "learning_rate": 1.2715793193700088e-05, + "loss": 0.0319, + "step": 46770 + }, + { + "epoch": 1.443916291129082, + "grad_norm": 0.36938267946243286, + "learning_rate": 1.2697437086054664e-05, + "loss": 0.0388, + "step": 46780 + }, + { + "epoch": 1.4442249521575405, + "grad_norm": 0.29883483052253723, + "learning_rate": 1.2679092309942937e-05, + "loss": 0.0327, + "step": 46790 + }, + { + "epoch": 1.4445336131859992, + "grad_norm": 0.3796534836292267, + "learning_rate": 1.266075887093755e-05, + "loss": 0.035, + "step": 46800 + }, + { + "epoch": 1.4448422742144578, + "grad_norm": 0.3091701865196228, + "learning_rate": 1.2642436774607757e-05, + "loss": 0.0374, + "step": 46810 + }, + { + "epoch": 1.4451509352429162, + "grad_norm": 0.37441399693489075, + "learning_rate": 1.2624126026519278e-05, + "loss": 0.0354, + "step": 46820 + }, + { + "epoch": 1.4454595962713748, + "grad_norm": 0.3402165472507477, + "learning_rate": 1.2605826632234474e-05, + "loss": 0.0325, + "step": 46830 + }, + { + "epoch": 1.4457682572998334, + "grad_norm": 0.32431337237358093, + "learning_rate": 1.2587538597312198e-05, + "loss": 0.0346, + "step": 46840 + }, + { + "epoch": 1.4460769183282918, + "grad_norm": 0.31948596239089966, + "learning_rate": 1.2569261927307884e-05, + "loss": 0.0361, + "step": 46850 + }, + { + "epoch": 1.4463855793567504, + "grad_norm": 0.37052807211875916, + "learning_rate": 1.2550996627773493e-05, + "loss": 0.0335, + "step": 46860 + }, + { + "epoch": 1.446694240385209, + "grad_norm": 0.38269591331481934, + "learning_rate": 1.2532742704257527e-05, + "loss": 0.0354, + "step": 46870 + }, + { + "epoch": 1.4470029014136676, + "grad_norm": 0.2819397747516632, + "learning_rate": 1.2514500162305087e-05, + "loss": 0.0358, + "step": 46880 + }, + { + "epoch": 1.447311562442126, + "grad_norm": 0.537837028503418, + "learning_rate": 1.2496269007457728e-05, + "loss": 0.0327, + "step": 46890 + }, + { + "epoch": 1.4476202234705846, + "grad_norm": 0.3588466942310333, + "learning_rate": 1.2478049245253625e-05, + "loss": 0.0405, + "step": 46900 + }, + { + "epoch": 1.4479288844990432, + "grad_norm": 0.4684883952140808, + "learning_rate": 1.2459840881227459e-05, + "loss": 0.0363, + "step": 46910 + }, + { + "epoch": 1.4482375455275016, + "grad_norm": 0.3004429340362549, + "learning_rate": 1.2441643920910435e-05, + "loss": 0.039, + "step": 46920 + }, + { + "epoch": 1.4485462065559602, + "grad_norm": 0.3568364679813385, + "learning_rate": 1.2423458369830322e-05, + "loss": 0.0335, + "step": 46930 + }, + { + "epoch": 1.4488548675844188, + "grad_norm": 0.4006308317184448, + "learning_rate": 1.2405284233511406e-05, + "loss": 0.0354, + "step": 46940 + }, + { + "epoch": 1.4491635286128774, + "grad_norm": 0.2897900342941284, + "learning_rate": 1.2387121517474487e-05, + "loss": 0.0385, + "step": 46950 + }, + { + "epoch": 1.4494721896413358, + "grad_norm": 0.2818731665611267, + "learning_rate": 1.2368970227236975e-05, + "loss": 0.0346, + "step": 46960 + }, + { + "epoch": 1.4497808506697945, + "grad_norm": 0.37566542625427246, + "learning_rate": 1.2350830368312688e-05, + "loss": 0.0338, + "step": 46970 + }, + { + "epoch": 1.4500895116982528, + "grad_norm": 0.6670220494270325, + "learning_rate": 1.2332701946212083e-05, + "loss": 0.0365, + "step": 46980 + }, + { + "epoch": 1.4503981727267115, + "grad_norm": 0.25527268648147583, + "learning_rate": 1.2314584966442077e-05, + "loss": 0.0409, + "step": 46990 + }, + { + "epoch": 1.45070683375517, + "grad_norm": 0.2833879888057709, + "learning_rate": 1.2296479434506136e-05, + "loss": 0.0365, + "step": 47000 + }, + { + "epoch": 1.4510154947836287, + "grad_norm": 0.31589969992637634, + "learning_rate": 1.2278385355904232e-05, + "loss": 0.0367, + "step": 47010 + }, + { + "epoch": 1.4513241558120873, + "grad_norm": 0.2863037586212158, + "learning_rate": 1.2260302736132867e-05, + "loss": 0.04, + "step": 47020 + }, + { + "epoch": 1.4516328168405457, + "grad_norm": 0.31096020340919495, + "learning_rate": 1.2242231580685098e-05, + "loss": 0.0366, + "step": 47030 + }, + { + "epoch": 1.4519414778690043, + "grad_norm": 0.32143986225128174, + "learning_rate": 1.2224171895050413e-05, + "loss": 0.0289, + "step": 47040 + }, + { + "epoch": 1.4522501388974627, + "grad_norm": 0.4260082244873047, + "learning_rate": 1.2206123684714903e-05, + "loss": 0.0339, + "step": 47050 + }, + { + "epoch": 1.4525587999259213, + "grad_norm": 0.3802686035633087, + "learning_rate": 1.2188086955161132e-05, + "loss": 0.0357, + "step": 47060 + }, + { + "epoch": 1.45286746095438, + "grad_norm": 0.41575154662132263, + "learning_rate": 1.2170061711868175e-05, + "loss": 0.0377, + "step": 47070 + }, + { + "epoch": 1.4531761219828385, + "grad_norm": 0.4507158696651459, + "learning_rate": 1.215204796031163e-05, + "loss": 0.0351, + "step": 47080 + }, + { + "epoch": 1.4534847830112971, + "grad_norm": 0.319987028837204, + "learning_rate": 1.2134045705963599e-05, + "loss": 0.0367, + "step": 47090 + }, + { + "epoch": 1.4537934440397555, + "grad_norm": 0.28611093759536743, + "learning_rate": 1.2116054954292689e-05, + "loss": 0.0331, + "step": 47100 + }, + { + "epoch": 1.4541021050682141, + "grad_norm": 0.34347420930862427, + "learning_rate": 1.2098075710764011e-05, + "loss": 0.0328, + "step": 47110 + }, + { + "epoch": 1.4544107660966725, + "grad_norm": 0.33531656861305237, + "learning_rate": 1.2080107980839183e-05, + "loss": 0.0392, + "step": 47120 + }, + { + "epoch": 1.4547194271251311, + "grad_norm": 0.2483757883310318, + "learning_rate": 1.2062151769976343e-05, + "loss": 0.0353, + "step": 47130 + }, + { + "epoch": 1.4550280881535897, + "grad_norm": 0.26541173458099365, + "learning_rate": 1.204420708363011e-05, + "loss": 0.035, + "step": 47140 + }, + { + "epoch": 1.4553367491820484, + "grad_norm": 0.3090918958187103, + "learning_rate": 1.2026273927251597e-05, + "loss": 0.035, + "step": 47150 + }, + { + "epoch": 1.4556454102105068, + "grad_norm": 0.2855260968208313, + "learning_rate": 1.2008352306288424e-05, + "loss": 0.0348, + "step": 47160 + }, + { + "epoch": 1.4559540712389654, + "grad_norm": 0.7842392921447754, + "learning_rate": 1.1990442226184695e-05, + "loss": 0.0384, + "step": 47170 + }, + { + "epoch": 1.456262732267424, + "grad_norm": 0.29448384046554565, + "learning_rate": 1.1972543692381066e-05, + "loss": 0.0364, + "step": 47180 + }, + { + "epoch": 1.4565713932958824, + "grad_norm": 0.34863147139549255, + "learning_rate": 1.1954656710314576e-05, + "loss": 0.0343, + "step": 47190 + }, + { + "epoch": 1.456880054324341, + "grad_norm": 0.31132641434669495, + "learning_rate": 1.1936781285418875e-05, + "loss": 0.0347, + "step": 47200 + }, + { + "epoch": 1.4571887153527996, + "grad_norm": 0.48597705364227295, + "learning_rate": 1.1918917423123993e-05, + "loss": 0.0359, + "step": 47210 + }, + { + "epoch": 1.4574973763812582, + "grad_norm": 0.3871159851551056, + "learning_rate": 1.1901065128856537e-05, + "loss": 0.0344, + "step": 47220 + }, + { + "epoch": 1.4578060374097166, + "grad_norm": 0.3943856358528137, + "learning_rate": 1.1883224408039551e-05, + "loss": 0.0386, + "step": 47230 + }, + { + "epoch": 1.4581146984381752, + "grad_norm": 0.32331427931785583, + "learning_rate": 1.1865395266092578e-05, + "loss": 0.0327, + "step": 47240 + }, + { + "epoch": 1.4584233594666338, + "grad_norm": 0.3452152609825134, + "learning_rate": 1.1847577708431633e-05, + "loss": 0.0348, + "step": 47250 + }, + { + "epoch": 1.4587320204950922, + "grad_norm": 0.4159036874771118, + "learning_rate": 1.1829771740469225e-05, + "loss": 0.0335, + "step": 47260 + }, + { + "epoch": 1.4590406815235508, + "grad_norm": 0.40712329745292664, + "learning_rate": 1.1811977367614324e-05, + "loss": 0.0408, + "step": 47270 + }, + { + "epoch": 1.4593493425520094, + "grad_norm": 0.23875093460083008, + "learning_rate": 1.1794194595272412e-05, + "loss": 0.034, + "step": 47280 + }, + { + "epoch": 1.459658003580468, + "grad_norm": 0.43997621536254883, + "learning_rate": 1.1776423428845423e-05, + "loss": 0.0358, + "step": 47290 + }, + { + "epoch": 1.4599666646089264, + "grad_norm": 0.3326110541820526, + "learning_rate": 1.1758663873731756e-05, + "loss": 0.0339, + "step": 47300 + }, + { + "epoch": 1.460275325637385, + "grad_norm": 0.3108561635017395, + "learning_rate": 1.1740915935326302e-05, + "loss": 0.0305, + "step": 47310 + }, + { + "epoch": 1.4605839866658434, + "grad_norm": 0.3554220199584961, + "learning_rate": 1.1723179619020396e-05, + "loss": 0.0342, + "step": 47320 + }, + { + "epoch": 1.460892647694302, + "grad_norm": 0.4778611660003662, + "learning_rate": 1.1705454930201914e-05, + "loss": 0.0354, + "step": 47330 + }, + { + "epoch": 1.4612013087227607, + "grad_norm": 0.30653998255729675, + "learning_rate": 1.1687741874255087e-05, + "loss": 0.0356, + "step": 47340 + }, + { + "epoch": 1.4615099697512193, + "grad_norm": 0.5343520641326904, + "learning_rate": 1.1670040456560728e-05, + "loss": 0.0367, + "step": 47350 + }, + { + "epoch": 1.4618186307796779, + "grad_norm": 0.2546285092830658, + "learning_rate": 1.1652350682496005e-05, + "loss": 0.0364, + "step": 47360 + }, + { + "epoch": 1.4621272918081363, + "grad_norm": 0.32557305693626404, + "learning_rate": 1.163467255743465e-05, + "loss": 0.0366, + "step": 47370 + }, + { + "epoch": 1.4624359528365949, + "grad_norm": 0.33763694763183594, + "learning_rate": 1.1617006086746796e-05, + "loss": 0.0368, + "step": 47380 + }, + { + "epoch": 1.4627446138650533, + "grad_norm": 0.3508402109146118, + "learning_rate": 1.1599351275799047e-05, + "loss": 0.0367, + "step": 47390 + }, + { + "epoch": 1.463053274893512, + "grad_norm": 0.34797340631484985, + "learning_rate": 1.1581708129954466e-05, + "loss": 0.0328, + "step": 47400 + }, + { + "epoch": 1.4633619359219705, + "grad_norm": 0.2957676649093628, + "learning_rate": 1.1564076654572587e-05, + "loss": 0.0356, + "step": 47410 + }, + { + "epoch": 1.4636705969504291, + "grad_norm": 0.28626155853271484, + "learning_rate": 1.1546456855009358e-05, + "loss": 0.0324, + "step": 47420 + }, + { + "epoch": 1.4639792579788875, + "grad_norm": 0.2875004708766937, + "learning_rate": 1.1528848736617248e-05, + "loss": 0.0347, + "step": 47430 + }, + { + "epoch": 1.4642879190073461, + "grad_norm": 0.3808160424232483, + "learning_rate": 1.1511252304745112e-05, + "loss": 0.0353, + "step": 47440 + }, + { + "epoch": 1.4645965800358047, + "grad_norm": 0.4169534146785736, + "learning_rate": 1.1493667564738297e-05, + "loss": 0.0334, + "step": 47450 + }, + { + "epoch": 1.4649052410642631, + "grad_norm": 0.35662347078323364, + "learning_rate": 1.1476094521938574e-05, + "loss": 0.0347, + "step": 47460 + }, + { + "epoch": 1.4652139020927217, + "grad_norm": 0.5833011865615845, + "learning_rate": 1.1458533181684167e-05, + "loss": 0.0353, + "step": 47470 + }, + { + "epoch": 1.4655225631211803, + "grad_norm": 0.3287070393562317, + "learning_rate": 1.1440983549309753e-05, + "loss": 0.0347, + "step": 47480 + }, + { + "epoch": 1.465831224149639, + "grad_norm": 0.4274834990501404, + "learning_rate": 1.1423445630146434e-05, + "loss": 0.0374, + "step": 47490 + }, + { + "epoch": 1.4661398851780973, + "grad_norm": 0.29322099685668945, + "learning_rate": 1.1405919429521799e-05, + "loss": 0.0342, + "step": 47500 + }, + { + "epoch": 1.466448546206556, + "grad_norm": 0.34371569752693176, + "learning_rate": 1.1388404952759802e-05, + "loss": 0.038, + "step": 47510 + }, + { + "epoch": 1.4667572072350146, + "grad_norm": 0.30463361740112305, + "learning_rate": 1.1370902205180923e-05, + "loss": 0.0358, + "step": 47520 + }, + { + "epoch": 1.467065868263473, + "grad_norm": 0.453630268573761, + "learning_rate": 1.1353411192101987e-05, + "loss": 0.0354, + "step": 47530 + }, + { + "epoch": 1.4673745292919316, + "grad_norm": 0.2941367030143738, + "learning_rate": 1.133593191883634e-05, + "loss": 0.0328, + "step": 47540 + }, + { + "epoch": 1.4676831903203902, + "grad_norm": 0.47927752137184143, + "learning_rate": 1.1318464390693711e-05, + "loss": 0.0399, + "step": 47550 + }, + { + "epoch": 1.4679918513488488, + "grad_norm": 0.36354950070381165, + "learning_rate": 1.1301008612980257e-05, + "loss": 0.0361, + "step": 47560 + }, + { + "epoch": 1.4683005123773072, + "grad_norm": 0.2848316729068756, + "learning_rate": 1.128356459099863e-05, + "loss": 0.0323, + "step": 47570 + }, + { + "epoch": 1.4686091734057658, + "grad_norm": 0.26921308040618896, + "learning_rate": 1.1266132330047802e-05, + "loss": 0.0408, + "step": 47580 + }, + { + "epoch": 1.4689178344342244, + "grad_norm": 0.2977297306060791, + "learning_rate": 1.1248711835423281e-05, + "loss": 0.0385, + "step": 47590 + }, + { + "epoch": 1.4692264954626828, + "grad_norm": 0.2839059829711914, + "learning_rate": 1.123130311241693e-05, + "loss": 0.032, + "step": 47600 + }, + { + "epoch": 1.4695351564911414, + "grad_norm": 0.2530215382575989, + "learning_rate": 1.1213906166317068e-05, + "loss": 0.0373, + "step": 47610 + }, + { + "epoch": 1.4698438175196, + "grad_norm": 0.2577265501022339, + "learning_rate": 1.1196521002408427e-05, + "loss": 0.0352, + "step": 47620 + }, + { + "epoch": 1.4701524785480586, + "grad_norm": 0.4613233804702759, + "learning_rate": 1.1179147625972159e-05, + "loss": 0.0399, + "step": 47630 + }, + { + "epoch": 1.470461139576517, + "grad_norm": 0.26425158977508545, + "learning_rate": 1.1161786042285822e-05, + "loss": 0.0353, + "step": 47640 + }, + { + "epoch": 1.4707698006049756, + "grad_norm": 0.36645838618278503, + "learning_rate": 1.1144436256623447e-05, + "loss": 0.0338, + "step": 47650 + }, + { + "epoch": 1.471078461633434, + "grad_norm": 0.35033097863197327, + "learning_rate": 1.1127098274255392e-05, + "loss": 0.034, + "step": 47660 + }, + { + "epoch": 1.4713871226618926, + "grad_norm": 0.29144713282585144, + "learning_rate": 1.1109772100448512e-05, + "loss": 0.0333, + "step": 47670 + }, + { + "epoch": 1.4716957836903513, + "grad_norm": 0.27113813161849976, + "learning_rate": 1.1092457740466033e-05, + "loss": 0.0328, + "step": 47680 + }, + { + "epoch": 1.4720044447188099, + "grad_norm": 0.48445063829421997, + "learning_rate": 1.10751551995676e-05, + "loss": 0.0355, + "step": 47690 + }, + { + "epoch": 1.4723131057472685, + "grad_norm": 0.3185565769672394, + "learning_rate": 1.1057864483009262e-05, + "loss": 0.0333, + "step": 47700 + }, + { + "epoch": 1.4726217667757269, + "grad_norm": 0.4730447232723236, + "learning_rate": 1.1040585596043473e-05, + "loss": 0.0348, + "step": 47710 + }, + { + "epoch": 1.4729304278041855, + "grad_norm": 0.5254570245742798, + "learning_rate": 1.1023318543919148e-05, + "loss": 0.035, + "step": 47720 + }, + { + "epoch": 1.4732390888326439, + "grad_norm": 0.28415417671203613, + "learning_rate": 1.10060633318815e-05, + "loss": 0.0393, + "step": 47730 + }, + { + "epoch": 1.4735477498611025, + "grad_norm": 0.38500580191612244, + "learning_rate": 1.0988819965172248e-05, + "loss": 0.0368, + "step": 47740 + }, + { + "epoch": 1.473856410889561, + "grad_norm": 0.3078926205635071, + "learning_rate": 1.0971588449029462e-05, + "loss": 0.0316, + "step": 47750 + }, + { + "epoch": 1.4741650719180197, + "grad_norm": 0.46886464953422546, + "learning_rate": 1.095436878868762e-05, + "loss": 0.0347, + "step": 47760 + }, + { + "epoch": 1.474473732946478, + "grad_norm": 0.36376330256462097, + "learning_rate": 1.0937160989377598e-05, + "loss": 0.0373, + "step": 47770 + }, + { + "epoch": 1.4747823939749367, + "grad_norm": 0.23339951038360596, + "learning_rate": 1.0919965056326676e-05, + "loss": 0.0375, + "step": 47780 + }, + { + "epoch": 1.4750910550033953, + "grad_norm": 0.32754361629486084, + "learning_rate": 1.0902780994758504e-05, + "loss": 0.0353, + "step": 47790 + }, + { + "epoch": 1.4753997160318537, + "grad_norm": 0.3397318124771118, + "learning_rate": 1.0885608809893193e-05, + "loss": 0.0326, + "step": 47800 + }, + { + "epoch": 1.4757083770603123, + "grad_norm": 0.5263281464576721, + "learning_rate": 1.0868448506947142e-05, + "loss": 0.0355, + "step": 47810 + }, + { + "epoch": 1.476017038088771, + "grad_norm": 0.32363319396972656, + "learning_rate": 1.0851300091133243e-05, + "loss": 0.0358, + "step": 47820 + }, + { + "epoch": 1.4763256991172296, + "grad_norm": 0.47704949975013733, + "learning_rate": 1.083416356766071e-05, + "loss": 0.0355, + "step": 47830 + }, + { + "epoch": 1.476634360145688, + "grad_norm": 0.4793188273906708, + "learning_rate": 1.0817038941735175e-05, + "loss": 0.035, + "step": 47840 + }, + { + "epoch": 1.4769430211741466, + "grad_norm": 0.2731672525405884, + "learning_rate": 1.0799926218558642e-05, + "loss": 0.0351, + "step": 47850 + }, + { + "epoch": 1.4772516822026052, + "grad_norm": 0.3456529378890991, + "learning_rate": 1.0782825403329488e-05, + "loss": 0.0348, + "step": 47860 + }, + { + "epoch": 1.4775603432310636, + "grad_norm": 0.39192306995391846, + "learning_rate": 1.076573650124254e-05, + "loss": 0.0359, + "step": 47870 + }, + { + "epoch": 1.4778690042595222, + "grad_norm": 0.394601970911026, + "learning_rate": 1.0748659517488891e-05, + "loss": 0.0342, + "step": 47880 + }, + { + "epoch": 1.4781776652879808, + "grad_norm": 0.2720644474029541, + "learning_rate": 1.0731594457256138e-05, + "loss": 0.0373, + "step": 47890 + }, + { + "epoch": 1.4784863263164394, + "grad_norm": 0.23486988246440887, + "learning_rate": 1.0714541325728139e-05, + "loss": 0.0395, + "step": 47900 + }, + { + "epoch": 1.4787949873448978, + "grad_norm": 0.2912793457508087, + "learning_rate": 1.0697500128085231e-05, + "loss": 0.0354, + "step": 47910 + }, + { + "epoch": 1.4791036483733564, + "grad_norm": 0.3705580532550812, + "learning_rate": 1.0680470869504055e-05, + "loss": 0.036, + "step": 47920 + }, + { + "epoch": 1.479412309401815, + "grad_norm": 0.2462342381477356, + "learning_rate": 1.066345355515766e-05, + "loss": 0.036, + "step": 47930 + }, + { + "epoch": 1.4797209704302734, + "grad_norm": 0.5561299324035645, + "learning_rate": 1.0646448190215453e-05, + "loss": 0.038, + "step": 47940 + }, + { + "epoch": 1.480029631458732, + "grad_norm": 0.3702816963195801, + "learning_rate": 1.0629454779843217e-05, + "loss": 0.0362, + "step": 47950 + }, + { + "epoch": 1.4803382924871906, + "grad_norm": 0.4526386857032776, + "learning_rate": 1.0612473329203082e-05, + "loss": 0.0386, + "step": 47960 + }, + { + "epoch": 1.4806469535156492, + "grad_norm": 0.2760521173477173, + "learning_rate": 1.0595503843453596e-05, + "loss": 0.035, + "step": 47970 + }, + { + "epoch": 1.4809556145441076, + "grad_norm": 0.3359192907810211, + "learning_rate": 1.0578546327749634e-05, + "loss": 0.0314, + "step": 47980 + }, + { + "epoch": 1.4812642755725662, + "grad_norm": 0.23150748014450073, + "learning_rate": 1.0561600787242425e-05, + "loss": 0.0333, + "step": 47990 + }, + { + "epoch": 1.4815729366010246, + "grad_norm": 0.2792832851409912, + "learning_rate": 1.0544667227079591e-05, + "loss": 0.0317, + "step": 48000 + }, + { + "epoch": 1.4818815976294832, + "grad_norm": 0.5733996033668518, + "learning_rate": 1.0527745652405085e-05, + "loss": 0.0321, + "step": 48010 + }, + { + "epoch": 1.4821902586579419, + "grad_norm": 0.3024829626083374, + "learning_rate": 1.051083606835927e-05, + "loss": 0.0361, + "step": 48020 + }, + { + "epoch": 1.4824989196864005, + "grad_norm": 0.253537118434906, + "learning_rate": 1.049393848007878e-05, + "loss": 0.0317, + "step": 48030 + }, + { + "epoch": 1.482807580714859, + "grad_norm": 0.5861140489578247, + "learning_rate": 1.0477052892696709e-05, + "loss": 0.0363, + "step": 48040 + }, + { + "epoch": 1.4831162417433175, + "grad_norm": 0.33929193019866943, + "learning_rate": 1.0460179311342394e-05, + "loss": 0.0311, + "step": 48050 + }, + { + "epoch": 1.483424902771776, + "grad_norm": 0.2547885477542877, + "learning_rate": 1.0443317741141634e-05, + "loss": 0.0405, + "step": 48060 + }, + { + "epoch": 1.4837335638002345, + "grad_norm": 0.3986351490020752, + "learning_rate": 1.0426468187216514e-05, + "loss": 0.0342, + "step": 48070 + }, + { + "epoch": 1.484042224828693, + "grad_norm": 0.27198469638824463, + "learning_rate": 1.0409630654685477e-05, + "loss": 0.0325, + "step": 48080 + }, + { + "epoch": 1.4843508858571517, + "grad_norm": 0.39720478653907776, + "learning_rate": 1.039280514866332e-05, + "loss": 0.0305, + "step": 48090 + }, + { + "epoch": 1.4846595468856103, + "grad_norm": 0.26622265577316284, + "learning_rate": 1.0375991674261198e-05, + "loss": 0.0352, + "step": 48100 + }, + { + "epoch": 1.4849682079140687, + "grad_norm": 0.3237873613834381, + "learning_rate": 1.0359190236586575e-05, + "loss": 0.0372, + "step": 48110 + }, + { + "epoch": 1.4852768689425273, + "grad_norm": 0.29899531602859497, + "learning_rate": 1.0342400840743322e-05, + "loss": 0.0351, + "step": 48120 + }, + { + "epoch": 1.485585529970986, + "grad_norm": 0.672107994556427, + "learning_rate": 1.0325623491831593e-05, + "loss": 0.0295, + "step": 48130 + }, + { + "epoch": 1.4858941909994443, + "grad_norm": 0.2686055600643158, + "learning_rate": 1.0308858194947906e-05, + "loss": 0.0383, + "step": 48140 + }, + { + "epoch": 1.486202852027903, + "grad_norm": 0.3501852750778198, + "learning_rate": 1.0292104955185111e-05, + "loss": 0.0346, + "step": 48150 + }, + { + "epoch": 1.4865115130563615, + "grad_norm": 0.3144979178905487, + "learning_rate": 1.0275363777632396e-05, + "loss": 0.0346, + "step": 48160 + }, + { + "epoch": 1.4868201740848201, + "grad_norm": 0.3780585825443268, + "learning_rate": 1.0258634667375321e-05, + "loss": 0.0373, + "step": 48170 + }, + { + "epoch": 1.4871288351132785, + "grad_norm": 0.4190988540649414, + "learning_rate": 1.02419176294957e-05, + "loss": 0.0358, + "step": 48180 + }, + { + "epoch": 1.4874374961417371, + "grad_norm": 0.27448809146881104, + "learning_rate": 1.0225212669071782e-05, + "loss": 0.0326, + "step": 48190 + }, + { + "epoch": 1.4877461571701958, + "grad_norm": 0.2821130156517029, + "learning_rate": 1.0208519791178029e-05, + "loss": 0.0303, + "step": 48200 + }, + { + "epoch": 1.4880548181986542, + "grad_norm": 0.25729313492774963, + "learning_rate": 1.019183900088535e-05, + "loss": 0.0351, + "step": 48210 + }, + { + "epoch": 1.4883634792271128, + "grad_norm": 0.380868136882782, + "learning_rate": 1.0175170303260906e-05, + "loss": 0.0324, + "step": 48220 + }, + { + "epoch": 1.4886721402555714, + "grad_norm": 0.3570501208305359, + "learning_rate": 1.0158513703368206e-05, + "loss": 0.0354, + "step": 48230 + }, + { + "epoch": 1.48898080128403, + "grad_norm": 0.32773885130882263, + "learning_rate": 1.0141869206267095e-05, + "loss": 0.0344, + "step": 48240 + }, + { + "epoch": 1.4892894623124884, + "grad_norm": 0.29869017004966736, + "learning_rate": 1.0125236817013723e-05, + "loss": 0.0383, + "step": 48250 + }, + { + "epoch": 1.489598123340947, + "grad_norm": 0.5280796885490417, + "learning_rate": 1.010861654066056e-05, + "loss": 0.0361, + "step": 48260 + }, + { + "epoch": 1.4899067843694056, + "grad_norm": 0.27932706475257874, + "learning_rate": 1.0092008382256434e-05, + "loss": 0.0363, + "step": 48270 + }, + { + "epoch": 1.490215445397864, + "grad_norm": 0.22967159748077393, + "learning_rate": 1.0075412346846458e-05, + "loss": 0.0331, + "step": 48280 + }, + { + "epoch": 1.4905241064263226, + "grad_norm": 0.3646487593650818, + "learning_rate": 1.0058828439472056e-05, + "loss": 0.0343, + "step": 48290 + }, + { + "epoch": 1.4908327674547812, + "grad_norm": 0.3917545676231384, + "learning_rate": 1.0042256665170996e-05, + "loss": 0.0361, + "step": 48300 + }, + { + "epoch": 1.4911414284832398, + "grad_norm": 0.4809851050376892, + "learning_rate": 1.0025697028977332e-05, + "loss": 0.0352, + "step": 48310 + }, + { + "epoch": 1.4914500895116982, + "grad_norm": 0.37029021978378296, + "learning_rate": 1.0009149535921454e-05, + "loss": 0.0343, + "step": 48320 + }, + { + "epoch": 1.4917587505401568, + "grad_norm": 0.31580790877342224, + "learning_rate": 9.992614191030031e-06, + "loss": 0.0336, + "step": 48330 + }, + { + "epoch": 1.4920674115686152, + "grad_norm": 0.3866753876209259, + "learning_rate": 9.976090999326115e-06, + "loss": 0.0313, + "step": 48340 + }, + { + "epoch": 1.4923760725970738, + "grad_norm": 0.4750453233718872, + "learning_rate": 9.959579965828952e-06, + "loss": 0.0366, + "step": 48350 + }, + { + "epoch": 1.4926847336255324, + "grad_norm": 0.32994794845581055, + "learning_rate": 9.943081095554218e-06, + "loss": 0.0347, + "step": 48360 + }, + { + "epoch": 1.492993394653991, + "grad_norm": 0.3240688741207123, + "learning_rate": 9.926594393513783e-06, + "loss": 0.035, + "step": 48370 + }, + { + "epoch": 1.4933020556824497, + "grad_norm": 0.33548909425735474, + "learning_rate": 9.910119864715906e-06, + "loss": 0.0344, + "step": 48380 + }, + { + "epoch": 1.493610716710908, + "grad_norm": 0.22933577001094818, + "learning_rate": 9.8936575141651e-06, + "loss": 0.0329, + "step": 48390 + }, + { + "epoch": 1.4939193777393667, + "grad_norm": 0.41471150517463684, + "learning_rate": 9.877207346862194e-06, + "loss": 0.033, + "step": 48400 + }, + { + "epoch": 1.494228038767825, + "grad_norm": 0.3922326862812042, + "learning_rate": 9.860769367804312e-06, + "loss": 0.0377, + "step": 48410 + }, + { + "epoch": 1.4945366997962837, + "grad_norm": 0.33007797598838806, + "learning_rate": 9.844343581984877e-06, + "loss": 0.0329, + "step": 48420 + }, + { + "epoch": 1.4948453608247423, + "grad_norm": 0.3918934762477875, + "learning_rate": 9.82792999439362e-06, + "loss": 0.0342, + "step": 48430 + }, + { + "epoch": 1.495154021853201, + "grad_norm": 0.4418620467185974, + "learning_rate": 9.811528610016546e-06, + "loss": 0.0366, + "step": 48440 + }, + { + "epoch": 1.4954626828816593, + "grad_norm": 0.3003976047039032, + "learning_rate": 9.79513943383597e-06, + "loss": 0.0413, + "step": 48450 + }, + { + "epoch": 1.495771343910118, + "grad_norm": 0.4561157822608948, + "learning_rate": 9.778762470830489e-06, + "loss": 0.0304, + "step": 48460 + }, + { + "epoch": 1.4960800049385765, + "grad_norm": 0.5031680464744568, + "learning_rate": 9.762397725974982e-06, + "loss": 0.0349, + "step": 48470 + }, + { + "epoch": 1.496388665967035, + "grad_norm": 0.4133344888687134, + "learning_rate": 9.746045204240622e-06, + "loss": 0.0335, + "step": 48480 + }, + { + "epoch": 1.4966973269954935, + "grad_norm": 0.36040425300598145, + "learning_rate": 9.729704910594917e-06, + "loss": 0.0327, + "step": 48490 + }, + { + "epoch": 1.4970059880239521, + "grad_norm": 0.3552595376968384, + "learning_rate": 9.713376850001554e-06, + "loss": 0.0371, + "step": 48500 + }, + { + "epoch": 1.4973146490524107, + "grad_norm": 0.2951916754245758, + "learning_rate": 9.697061027420622e-06, + "loss": 0.0356, + "step": 48510 + }, + { + "epoch": 1.4976233100808691, + "grad_norm": 0.3974596858024597, + "learning_rate": 9.680757447808385e-06, + "loss": 0.0367, + "step": 48520 + }, + { + "epoch": 1.4979319711093277, + "grad_norm": 0.329119473695755, + "learning_rate": 9.664466116117488e-06, + "loss": 0.0377, + "step": 48530 + }, + { + "epoch": 1.4982406321377864, + "grad_norm": 0.45495861768722534, + "learning_rate": 9.64818703729678e-06, + "loss": 0.0332, + "step": 48540 + }, + { + "epoch": 1.4985492931662447, + "grad_norm": 0.2595498263835907, + "learning_rate": 9.631920216291423e-06, + "loss": 0.0377, + "step": 48550 + }, + { + "epoch": 1.4988579541947034, + "grad_norm": 0.4566974639892578, + "learning_rate": 9.615665658042849e-06, + "loss": 0.0417, + "step": 48560 + }, + { + "epoch": 1.499166615223162, + "grad_norm": 0.3299633860588074, + "learning_rate": 9.599423367488747e-06, + "loss": 0.0338, + "step": 48570 + }, + { + "epoch": 1.4994752762516206, + "grad_norm": 0.3263728618621826, + "learning_rate": 9.583193349563124e-06, + "loss": 0.0357, + "step": 48580 + }, + { + "epoch": 1.499783937280079, + "grad_norm": 0.3338775932788849, + "learning_rate": 9.566975609196216e-06, + "loss": 0.0365, + "step": 48590 + }, + { + "epoch": 1.5000925983085376, + "grad_norm": 0.42016690969467163, + "learning_rate": 9.550770151314548e-06, + "loss": 0.0395, + "step": 48600 + }, + { + "epoch": 1.500401259336996, + "grad_norm": 0.41950172185897827, + "learning_rate": 9.53457698084091e-06, + "loss": 0.0358, + "step": 48610 + }, + { + "epoch": 1.5007099203654546, + "grad_norm": 0.2741645276546478, + "learning_rate": 9.518396102694355e-06, + "loss": 0.0381, + "step": 48620 + }, + { + "epoch": 1.5010185813939132, + "grad_norm": 0.34390804171562195, + "learning_rate": 9.502227521790198e-06, + "loss": 0.0379, + "step": 48630 + }, + { + "epoch": 1.5013272424223718, + "grad_norm": 0.34637221693992615, + "learning_rate": 9.486071243040063e-06, + "loss": 0.0339, + "step": 48640 + }, + { + "epoch": 1.5016359034508304, + "grad_norm": 0.4828738868236542, + "learning_rate": 9.469927271351747e-06, + "loss": 0.0367, + "step": 48650 + }, + { + "epoch": 1.5019445644792888, + "grad_norm": 0.46810007095336914, + "learning_rate": 9.453795611629419e-06, + "loss": 0.0364, + "step": 48660 + }, + { + "epoch": 1.5022532255077474, + "grad_norm": 0.33315303921699524, + "learning_rate": 9.437676268773399e-06, + "loss": 0.0321, + "step": 48670 + }, + { + "epoch": 1.5025618865362058, + "grad_norm": 0.33858540654182434, + "learning_rate": 9.421569247680357e-06, + "loss": 0.0332, + "step": 48680 + }, + { + "epoch": 1.5028705475646644, + "grad_norm": 0.3451688587665558, + "learning_rate": 9.40547455324316e-06, + "loss": 0.0338, + "step": 48690 + }, + { + "epoch": 1.503179208593123, + "grad_norm": 0.4742872416973114, + "learning_rate": 9.389392190350965e-06, + "loss": 0.0358, + "step": 48700 + }, + { + "epoch": 1.5034878696215817, + "grad_norm": 0.3289341926574707, + "learning_rate": 9.373322163889153e-06, + "loss": 0.0356, + "step": 48710 + }, + { + "epoch": 1.5037965306500403, + "grad_norm": 0.22905269265174866, + "learning_rate": 9.357264478739375e-06, + "loss": 0.0322, + "step": 48720 + }, + { + "epoch": 1.5041051916784987, + "grad_norm": 0.29535698890686035, + "learning_rate": 9.341219139779567e-06, + "loss": 0.0365, + "step": 48730 + }, + { + "epoch": 1.5044138527069573, + "grad_norm": 0.37525084614753723, + "learning_rate": 9.325186151883824e-06, + "loss": 0.0339, + "step": 48740 + }, + { + "epoch": 1.5047225137354157, + "grad_norm": 0.28682559728622437, + "learning_rate": 9.30916551992258e-06, + "loss": 0.0372, + "step": 48750 + }, + { + "epoch": 1.5050311747638743, + "grad_norm": 0.32288768887519836, + "learning_rate": 9.293157248762479e-06, + "loss": 0.0378, + "step": 48760 + }, + { + "epoch": 1.5053398357923329, + "grad_norm": 0.2929612398147583, + "learning_rate": 9.2771613432664e-06, + "loss": 0.0362, + "step": 48770 + }, + { + "epoch": 1.5056484968207915, + "grad_norm": 0.37327316403388977, + "learning_rate": 9.261177808293481e-06, + "loss": 0.0389, + "step": 48780 + }, + { + "epoch": 1.50595715784925, + "grad_norm": 0.312939316034317, + "learning_rate": 9.245206648699096e-06, + "loss": 0.0384, + "step": 48790 + }, + { + "epoch": 1.5062658188777085, + "grad_norm": 0.4200822710990906, + "learning_rate": 9.22924786933485e-06, + "loss": 0.041, + "step": 48800 + }, + { + "epoch": 1.5065744799061669, + "grad_norm": 0.2790292799472809, + "learning_rate": 9.213301475048642e-06, + "loss": 0.0351, + "step": 48810 + }, + { + "epoch": 1.5068831409346255, + "grad_norm": 0.2840653955936432, + "learning_rate": 9.197367470684504e-06, + "loss": 0.0322, + "step": 48820 + }, + { + "epoch": 1.507191801963084, + "grad_norm": 0.42191463708877563, + "learning_rate": 9.181445861082816e-06, + "loss": 0.0355, + "step": 48830 + }, + { + "epoch": 1.5075004629915427, + "grad_norm": 0.3948734700679779, + "learning_rate": 9.16553665108012e-06, + "loss": 0.0343, + "step": 48840 + }, + { + "epoch": 1.5078091240200013, + "grad_norm": 0.3687100112438202, + "learning_rate": 9.149639845509223e-06, + "loss": 0.0337, + "step": 48850 + }, + { + "epoch": 1.50811778504846, + "grad_norm": 0.2849563956260681, + "learning_rate": 9.133755449199144e-06, + "loss": 0.0359, + "step": 48860 + }, + { + "epoch": 1.5084264460769183, + "grad_norm": 0.32283568382263184, + "learning_rate": 9.117883466975135e-06, + "loss": 0.0302, + "step": 48870 + }, + { + "epoch": 1.5087351071053767, + "grad_norm": 0.24561633169651031, + "learning_rate": 9.10202390365873e-06, + "loss": 0.0346, + "step": 48880 + }, + { + "epoch": 1.5090437681338353, + "grad_norm": 0.376144140958786, + "learning_rate": 9.086176764067583e-06, + "loss": 0.0328, + "step": 48890 + }, + { + "epoch": 1.509352429162294, + "grad_norm": 0.3428276777267456, + "learning_rate": 9.070342053015684e-06, + "loss": 0.0347, + "step": 48900 + }, + { + "epoch": 1.5096610901907526, + "grad_norm": 0.3314542770385742, + "learning_rate": 9.054519775313187e-06, + "loss": 0.0302, + "step": 48910 + }, + { + "epoch": 1.5099697512192112, + "grad_norm": 0.3076273500919342, + "learning_rate": 9.038709935766476e-06, + "loss": 0.0323, + "step": 48920 + }, + { + "epoch": 1.5102784122476696, + "grad_norm": 0.3124814033508301, + "learning_rate": 9.02291253917817e-06, + "loss": 0.039, + "step": 48930 + }, + { + "epoch": 1.5105870732761282, + "grad_norm": 0.28928133845329285, + "learning_rate": 9.007127590347091e-06, + "loss": 0.0354, + "step": 48940 + }, + { + "epoch": 1.5108957343045866, + "grad_norm": 0.32306668162345886, + "learning_rate": 8.991355094068288e-06, + "loss": 0.0343, + "step": 48950 + }, + { + "epoch": 1.5112043953330452, + "grad_norm": 0.3333916962146759, + "learning_rate": 8.975595055133062e-06, + "loss": 0.0364, + "step": 48960 + }, + { + "epoch": 1.5115130563615038, + "grad_norm": 0.4153554141521454, + "learning_rate": 8.959847478328848e-06, + "loss": 0.0306, + "step": 48970 + }, + { + "epoch": 1.5118217173899624, + "grad_norm": 0.41617506742477417, + "learning_rate": 8.944112368439378e-06, + "loss": 0.0386, + "step": 48980 + }, + { + "epoch": 1.512130378418421, + "grad_norm": 0.2932833433151245, + "learning_rate": 8.928389730244552e-06, + "loss": 0.0354, + "step": 48990 + }, + { + "epoch": 1.5124390394468794, + "grad_norm": 0.3103211224079132, + "learning_rate": 8.912679568520494e-06, + "loss": 0.0357, + "step": 49000 + }, + { + "epoch": 1.512747700475338, + "grad_norm": 0.27120915055274963, + "learning_rate": 8.896981888039534e-06, + "loss": 0.0369, + "step": 49010 + }, + { + "epoch": 1.5130563615037964, + "grad_norm": 0.41633009910583496, + "learning_rate": 8.881296693570201e-06, + "loss": 0.0412, + "step": 49020 + }, + { + "epoch": 1.513365022532255, + "grad_norm": 0.2649095952510834, + "learning_rate": 8.865623989877281e-06, + "loss": 0.0326, + "step": 49030 + }, + { + "epoch": 1.5136736835607136, + "grad_norm": 0.29777127504348755, + "learning_rate": 8.849963781721681e-06, + "loss": 0.032, + "step": 49040 + }, + { + "epoch": 1.5139823445891722, + "grad_norm": 0.29972586035728455, + "learning_rate": 8.834316073860588e-06, + "loss": 0.032, + "step": 49050 + }, + { + "epoch": 1.5142910056176309, + "grad_norm": 0.2766883075237274, + "learning_rate": 8.818680871047357e-06, + "loss": 0.0366, + "step": 49060 + }, + { + "epoch": 1.5145996666460892, + "grad_norm": 0.22714544832706451, + "learning_rate": 8.803058178031549e-06, + "loss": 0.0332, + "step": 49070 + }, + { + "epoch": 1.5149083276745479, + "grad_norm": 0.2366216778755188, + "learning_rate": 8.787447999558922e-06, + "loss": 0.0355, + "step": 49080 + }, + { + "epoch": 1.5152169887030063, + "grad_norm": 0.30101442337036133, + "learning_rate": 8.77185034037144e-06, + "loss": 0.0354, + "step": 49090 + }, + { + "epoch": 1.5155256497314649, + "grad_norm": 0.5576549172401428, + "learning_rate": 8.756265205207259e-06, + "loss": 0.0368, + "step": 49100 + }, + { + "epoch": 1.5158343107599235, + "grad_norm": 0.3889298737049103, + "learning_rate": 8.740692598800732e-06, + "loss": 0.0377, + "step": 49110 + }, + { + "epoch": 1.516142971788382, + "grad_norm": 0.49573343992233276, + "learning_rate": 8.72513252588239e-06, + "loss": 0.0348, + "step": 49120 + }, + { + "epoch": 1.5164516328168407, + "grad_norm": 0.24688324332237244, + "learning_rate": 8.709584991178998e-06, + "loss": 0.0332, + "step": 49130 + }, + { + "epoch": 1.516760293845299, + "grad_norm": 0.32708311080932617, + "learning_rate": 8.694049999413479e-06, + "loss": 0.0322, + "step": 49140 + }, + { + "epoch": 1.5170689548737575, + "grad_norm": 0.3468136787414551, + "learning_rate": 8.678527555304945e-06, + "loss": 0.0383, + "step": 49150 + }, + { + "epoch": 1.517377615902216, + "grad_norm": 0.45878326892852783, + "learning_rate": 8.663017663568712e-06, + "loss": 0.0368, + "step": 49160 + }, + { + "epoch": 1.5176862769306747, + "grad_norm": 0.43139156699180603, + "learning_rate": 8.647520328916259e-06, + "loss": 0.0325, + "step": 49170 + }, + { + "epoch": 1.5179949379591333, + "grad_norm": 0.36854204535484314, + "learning_rate": 8.632035556055307e-06, + "loss": 0.037, + "step": 49180 + }, + { + "epoch": 1.518303598987592, + "grad_norm": 0.3553614318370819, + "learning_rate": 8.616563349689672e-06, + "loss": 0.0333, + "step": 49190 + }, + { + "epoch": 1.5186122600160503, + "grad_norm": 0.8517873883247375, + "learning_rate": 8.601103714519448e-06, + "loss": 0.0385, + "step": 49200 + }, + { + "epoch": 1.518920921044509, + "grad_norm": 0.29389864206314087, + "learning_rate": 8.58565665524082e-06, + "loss": 0.036, + "step": 49210 + }, + { + "epoch": 1.5192295820729673, + "grad_norm": 0.3676852583885193, + "learning_rate": 8.570222176546222e-06, + "loss": 0.0296, + "step": 49220 + }, + { + "epoch": 1.519538243101426, + "grad_norm": 0.43538519740104675, + "learning_rate": 8.554800283124242e-06, + "loss": 0.0386, + "step": 49230 + }, + { + "epoch": 1.5198469041298845, + "grad_norm": 0.2894030511379242, + "learning_rate": 8.539390979659639e-06, + "loss": 0.0383, + "step": 49240 + }, + { + "epoch": 1.5201555651583432, + "grad_norm": 0.6836541295051575, + "learning_rate": 8.523994270833352e-06, + "loss": 0.032, + "step": 49250 + }, + { + "epoch": 1.5204642261868018, + "grad_norm": 0.5142187476158142, + "learning_rate": 8.5086101613225e-06, + "loss": 0.033, + "step": 49260 + }, + { + "epoch": 1.5207728872152602, + "grad_norm": 0.4340974986553192, + "learning_rate": 8.493238655800346e-06, + "loss": 0.0396, + "step": 49270 + }, + { + "epoch": 1.5210815482437188, + "grad_norm": 0.23849543929100037, + "learning_rate": 8.47787975893638e-06, + "loss": 0.0329, + "step": 49280 + }, + { + "epoch": 1.5213902092721772, + "grad_norm": 0.3972793519496918, + "learning_rate": 8.462533475396211e-06, + "loss": 0.0369, + "step": 49290 + }, + { + "epoch": 1.5216988703006358, + "grad_norm": 0.45433005690574646, + "learning_rate": 8.447199809841643e-06, + "loss": 0.0322, + "step": 49300 + }, + { + "epoch": 1.5220075313290944, + "grad_norm": 0.3330115079879761, + "learning_rate": 8.431878766930635e-06, + "loss": 0.0317, + "step": 49310 + }, + { + "epoch": 1.522316192357553, + "grad_norm": 0.43722862005233765, + "learning_rate": 8.416570351317304e-06, + "loss": 0.0317, + "step": 49320 + }, + { + "epoch": 1.5226248533860116, + "grad_norm": 0.2968553304672241, + "learning_rate": 8.401274567651973e-06, + "loss": 0.0348, + "step": 49330 + }, + { + "epoch": 1.52293351441447, + "grad_norm": 0.6502068042755127, + "learning_rate": 8.385991420581058e-06, + "loss": 0.0364, + "step": 49340 + }, + { + "epoch": 1.5232421754429286, + "grad_norm": 0.328908234834671, + "learning_rate": 8.370720914747215e-06, + "loss": 0.0354, + "step": 49350 + }, + { + "epoch": 1.523550836471387, + "grad_norm": 0.46006205677986145, + "learning_rate": 8.355463054789181e-06, + "loss": 0.036, + "step": 49360 + }, + { + "epoch": 1.5238594974998456, + "grad_norm": 0.3006315529346466, + "learning_rate": 8.340217845341919e-06, + "loss": 0.0342, + "step": 49370 + }, + { + "epoch": 1.5241681585283042, + "grad_norm": 0.27883970737457275, + "learning_rate": 8.324985291036514e-06, + "loss": 0.035, + "step": 49380 + }, + { + "epoch": 1.5244768195567628, + "grad_norm": 0.4310315251350403, + "learning_rate": 8.309765396500213e-06, + "loss": 0.0317, + "step": 49390 + }, + { + "epoch": 1.5247854805852215, + "grad_norm": 0.43290776014328003, + "learning_rate": 8.294558166356419e-06, + "loss": 0.0326, + "step": 49400 + }, + { + "epoch": 1.5250941416136798, + "grad_norm": 0.2841067314147949, + "learning_rate": 8.279363605224683e-06, + "loss": 0.0422, + "step": 49410 + }, + { + "epoch": 1.5254028026421385, + "grad_norm": 0.5094163417816162, + "learning_rate": 8.264181717720704e-06, + "loss": 0.0363, + "step": 49420 + }, + { + "epoch": 1.5257114636705968, + "grad_norm": 0.4896634519100189, + "learning_rate": 8.249012508456361e-06, + "loss": 0.0371, + "step": 49430 + }, + { + "epoch": 1.5260201246990555, + "grad_norm": 0.26327410340309143, + "learning_rate": 8.233855982039646e-06, + "loss": 0.0376, + "step": 49440 + }, + { + "epoch": 1.526328785727514, + "grad_norm": 0.6621829867362976, + "learning_rate": 8.218712143074708e-06, + "loss": 0.0351, + "step": 49450 + }, + { + "epoch": 1.5266374467559727, + "grad_norm": 0.33004164695739746, + "learning_rate": 8.203580996161858e-06, + "loss": 0.0321, + "step": 49460 + }, + { + "epoch": 1.5269461077844313, + "grad_norm": 0.277147501707077, + "learning_rate": 8.188462545897512e-06, + "loss": 0.0323, + "step": 49470 + }, + { + "epoch": 1.5272547688128897, + "grad_norm": 0.4777902364730835, + "learning_rate": 8.173356796874304e-06, + "loss": 0.0379, + "step": 49480 + }, + { + "epoch": 1.527563429841348, + "grad_norm": 0.3653133809566498, + "learning_rate": 8.158263753680906e-06, + "loss": 0.0361, + "step": 49490 + }, + { + "epoch": 1.5278720908698067, + "grad_norm": 0.2741551399230957, + "learning_rate": 8.143183420902239e-06, + "loss": 0.0335, + "step": 49500 + }, + { + "epoch": 1.5281807518982653, + "grad_norm": 0.4758780896663666, + "learning_rate": 8.128115803119258e-06, + "loss": 0.0326, + "step": 49510 + }, + { + "epoch": 1.528489412926724, + "grad_norm": 0.4359132945537567, + "learning_rate": 8.11306090490916e-06, + "loss": 0.0348, + "step": 49520 + }, + { + "epoch": 1.5287980739551825, + "grad_norm": 0.2993414103984833, + "learning_rate": 8.098018730845169e-06, + "loss": 0.0319, + "step": 49530 + }, + { + "epoch": 1.529106734983641, + "grad_norm": 0.4581685960292816, + "learning_rate": 8.082989285496745e-06, + "loss": 0.0349, + "step": 49540 + }, + { + "epoch": 1.5294153960120995, + "grad_norm": 0.2624550461769104, + "learning_rate": 8.067972573429416e-06, + "loss": 0.0382, + "step": 49550 + }, + { + "epoch": 1.529724057040558, + "grad_norm": 0.3730414807796478, + "learning_rate": 8.052968599204874e-06, + "loss": 0.035, + "step": 49560 + }, + { + "epoch": 1.5300327180690165, + "grad_norm": 0.32728061079978943, + "learning_rate": 8.037977367380922e-06, + "loss": 0.0321, + "step": 49570 + }, + { + "epoch": 1.5303413790974751, + "grad_norm": 0.4011220932006836, + "learning_rate": 8.022998882511495e-06, + "loss": 0.0354, + "step": 49580 + }, + { + "epoch": 1.5306500401259338, + "grad_norm": 0.23280666768550873, + "learning_rate": 8.008033149146677e-06, + "loss": 0.0364, + "step": 49590 + }, + { + "epoch": 1.5309587011543924, + "grad_norm": 0.2944525480270386, + "learning_rate": 7.993080171832656e-06, + "loss": 0.0333, + "step": 49600 + }, + { + "epoch": 1.5312673621828508, + "grad_norm": 0.373310387134552, + "learning_rate": 7.978139955111752e-06, + "loss": 0.036, + "step": 49610 + }, + { + "epoch": 1.5315760232113094, + "grad_norm": 0.38389453291893005, + "learning_rate": 7.9632125035224e-06, + "loss": 0.0322, + "step": 49620 + }, + { + "epoch": 1.5318846842397678, + "grad_norm": 0.28570544719696045, + "learning_rate": 7.948297821599177e-06, + "loss": 0.0367, + "step": 49630 + }, + { + "epoch": 1.5321933452682264, + "grad_norm": 0.3175638020038605, + "learning_rate": 7.933395913872755e-06, + "loss": 0.0361, + "step": 49640 + }, + { + "epoch": 1.532502006296685, + "grad_norm": 0.41374877095222473, + "learning_rate": 7.918506784869972e-06, + "loss": 0.0337, + "step": 49650 + }, + { + "epoch": 1.5328106673251436, + "grad_norm": 0.30662113428115845, + "learning_rate": 7.903630439113707e-06, + "loss": 0.0334, + "step": 49660 + }, + { + "epoch": 1.5331193283536022, + "grad_norm": 0.3060745596885681, + "learning_rate": 7.888766881123044e-06, + "loss": 0.0329, + "step": 49670 + }, + { + "epoch": 1.5334279893820606, + "grad_norm": 0.34012046456336975, + "learning_rate": 7.873916115413099e-06, + "loss": 0.0336, + "step": 49680 + }, + { + "epoch": 1.5337366504105192, + "grad_norm": 0.3600308895111084, + "learning_rate": 7.85907814649518e-06, + "loss": 0.0362, + "step": 49690 + }, + { + "epoch": 1.5340453114389776, + "grad_norm": 0.3005325198173523, + "learning_rate": 7.844252978876649e-06, + "loss": 0.0343, + "step": 49700 + }, + { + "epoch": 1.5343539724674362, + "grad_norm": 1.361586332321167, + "learning_rate": 7.829440617061001e-06, + "loss": 0.0333, + "step": 49710 + }, + { + "epoch": 1.5346626334958948, + "grad_norm": 0.2736550271511078, + "learning_rate": 7.814641065547851e-06, + "loss": 0.0337, + "step": 49720 + }, + { + "epoch": 1.5349712945243534, + "grad_norm": 0.37539970874786377, + "learning_rate": 7.79985432883289e-06, + "loss": 0.0357, + "step": 49730 + }, + { + "epoch": 1.535279955552812, + "grad_norm": 0.36426493525505066, + "learning_rate": 7.78508041140797e-06, + "loss": 0.0325, + "step": 49740 + }, + { + "epoch": 1.5355886165812704, + "grad_norm": 0.941225528717041, + "learning_rate": 7.770319317760993e-06, + "loss": 0.0384, + "step": 49750 + }, + { + "epoch": 1.535897277609729, + "grad_norm": 0.34485694766044617, + "learning_rate": 7.755571052376004e-06, + "loss": 0.034, + "step": 49760 + }, + { + "epoch": 1.5362059386381874, + "grad_norm": 0.3894895911216736, + "learning_rate": 7.740835619733128e-06, + "loss": 0.0313, + "step": 49770 + }, + { + "epoch": 1.536514599666646, + "grad_norm": 0.2553911805152893, + "learning_rate": 7.726113024308601e-06, + "loss": 0.0349, + "step": 49780 + }, + { + "epoch": 1.5368232606951047, + "grad_norm": 0.3273946940898895, + "learning_rate": 7.711403270574746e-06, + "loss": 0.0334, + "step": 49790 + }, + { + "epoch": 1.5371319217235633, + "grad_norm": 0.32567596435546875, + "learning_rate": 7.696706363000039e-06, + "loss": 0.0339, + "step": 49800 + }, + { + "epoch": 1.537440582752022, + "grad_norm": 0.6439789533615112, + "learning_rate": 7.682022306048959e-06, + "loss": 0.0345, + "step": 49810 + }, + { + "epoch": 1.5377492437804803, + "grad_norm": 0.27249521017074585, + "learning_rate": 7.667351104182186e-06, + "loss": 0.0361, + "step": 49820 + }, + { + "epoch": 1.5380579048089387, + "grad_norm": 0.28467339277267456, + "learning_rate": 7.652692761856395e-06, + "loss": 0.0346, + "step": 49830 + }, + { + "epoch": 1.5383665658373973, + "grad_norm": 0.3349708318710327, + "learning_rate": 7.63804728352444e-06, + "loss": 0.0336, + "step": 49840 + }, + { + "epoch": 1.538675226865856, + "grad_norm": 0.2663988173007965, + "learning_rate": 7.623414673635215e-06, + "loss": 0.0333, + "step": 49850 + }, + { + "epoch": 1.5389838878943145, + "grad_norm": 0.27697890996932983, + "learning_rate": 7.608794936633723e-06, + "loss": 0.0371, + "step": 49860 + }, + { + "epoch": 1.5392925489227731, + "grad_norm": 0.37900862097740173, + "learning_rate": 7.594188076961056e-06, + "loss": 0.0359, + "step": 49870 + }, + { + "epoch": 1.5396012099512315, + "grad_norm": 0.26409661769866943, + "learning_rate": 7.579594099054382e-06, + "loss": 0.0336, + "step": 49880 + }, + { + "epoch": 1.5399098709796901, + "grad_norm": 0.34122249484062195, + "learning_rate": 7.565013007346983e-06, + "loss": 0.0365, + "step": 49890 + }, + { + "epoch": 1.5402185320081485, + "grad_norm": 0.44609546661376953, + "learning_rate": 7.5504448062682035e-06, + "loss": 0.0332, + "step": 49900 + }, + { + "epoch": 1.5405271930366071, + "grad_norm": 0.3409268260002136, + "learning_rate": 7.53588950024347e-06, + "loss": 0.0342, + "step": 49910 + }, + { + "epoch": 1.5408358540650657, + "grad_norm": 0.7941582202911377, + "learning_rate": 7.5213470936943145e-06, + "loss": 0.0452, + "step": 49920 + }, + { + "epoch": 1.5411445150935243, + "grad_norm": 0.40343332290649414, + "learning_rate": 7.506817591038323e-06, + "loss": 0.0303, + "step": 49930 + }, + { + "epoch": 1.541453176121983, + "grad_norm": 0.3845701813697815, + "learning_rate": 7.492300996689183e-06, + "loss": 0.0383, + "step": 49940 + }, + { + "epoch": 1.5417618371504413, + "grad_norm": 0.3963932394981384, + "learning_rate": 7.477797315056645e-06, + "loss": 0.0371, + "step": 49950 + }, + { + "epoch": 1.5420704981789, + "grad_norm": 0.2856624722480774, + "learning_rate": 7.463306550546539e-06, + "loss": 0.0325, + "step": 49960 + }, + { + "epoch": 1.5423791592073584, + "grad_norm": 0.7538663744926453, + "learning_rate": 7.448828707560812e-06, + "loss": 0.0326, + "step": 49970 + }, + { + "epoch": 1.542687820235817, + "grad_norm": 0.35511258244514465, + "learning_rate": 7.4343637904974e-06, + "loss": 0.0348, + "step": 49980 + }, + { + "epoch": 1.5429964812642756, + "grad_norm": 0.7348798513412476, + "learning_rate": 7.419911803750401e-06, + "loss": 0.036, + "step": 49990 + }, + { + "epoch": 1.5433051422927342, + "grad_norm": 0.3318471908569336, + "learning_rate": 7.405472751709935e-06, + "loss": 0.0315, + "step": 50000 + }, + { + "epoch": 1.5436138033211928, + "grad_norm": 0.3110131621360779, + "learning_rate": 7.3910466387622e-06, + "loss": 0.0352, + "step": 50010 + }, + { + "epoch": 1.5439224643496512, + "grad_norm": 0.30169960856437683, + "learning_rate": 7.3766334692894735e-06, + "loss": 0.0349, + "step": 50020 + }, + { + "epoch": 1.5442311253781098, + "grad_norm": 0.3786137104034424, + "learning_rate": 7.3622332476700865e-06, + "loss": 0.0302, + "step": 50030 + }, + { + "epoch": 1.5445397864065682, + "grad_norm": 0.3258340358734131, + "learning_rate": 7.347845978278472e-06, + "loss": 0.0355, + "step": 50040 + }, + { + "epoch": 1.5448484474350268, + "grad_norm": 0.3784841299057007, + "learning_rate": 7.333471665485065e-06, + "loss": 0.0355, + "step": 50050 + }, + { + "epoch": 1.5451571084634854, + "grad_norm": 0.3617163598537445, + "learning_rate": 7.31911031365643e-06, + "loss": 0.0371, + "step": 50060 + }, + { + "epoch": 1.545465769491944, + "grad_norm": 0.4274131953716278, + "learning_rate": 7.304761927155157e-06, + "loss": 0.0374, + "step": 50070 + }, + { + "epoch": 1.5457744305204026, + "grad_norm": 0.3738037049770355, + "learning_rate": 7.29042651033991e-06, + "loss": 0.0359, + "step": 50080 + }, + { + "epoch": 1.546083091548861, + "grad_norm": 0.3922238349914551, + "learning_rate": 7.276104067565409e-06, + "loss": 0.0346, + "step": 50090 + }, + { + "epoch": 1.5463917525773194, + "grad_norm": 0.37832584977149963, + "learning_rate": 7.261794603182431e-06, + "loss": 0.0358, + "step": 50100 + }, + { + "epoch": 1.546700413605778, + "grad_norm": 0.20556120574474335, + "learning_rate": 7.24749812153781e-06, + "loss": 0.0336, + "step": 50110 + }, + { + "epoch": 1.5470090746342366, + "grad_norm": 0.31419000029563904, + "learning_rate": 7.2332146269744605e-06, + "loss": 0.0341, + "step": 50120 + }, + { + "epoch": 1.5473177356626953, + "grad_norm": 0.32824191451072693, + "learning_rate": 7.218944123831295e-06, + "loss": 0.0347, + "step": 50130 + }, + { + "epoch": 1.5476263966911539, + "grad_norm": 0.3820224404335022, + "learning_rate": 7.204686616443351e-06, + "loss": 0.0331, + "step": 50140 + }, + { + "epoch": 1.5479350577196125, + "grad_norm": 0.37572425603866577, + "learning_rate": 7.190442109141665e-06, + "loss": 0.0396, + "step": 50150 + }, + { + "epoch": 1.5482437187480709, + "grad_norm": 0.3789660632610321, + "learning_rate": 7.176210606253347e-06, + "loss": 0.0316, + "step": 50160 + }, + { + "epoch": 1.5485523797765293, + "grad_norm": 0.4553125500679016, + "learning_rate": 7.161992112101551e-06, + "loss": 0.0348, + "step": 50170 + }, + { + "epoch": 1.5488610408049879, + "grad_norm": 0.5181511044502258, + "learning_rate": 7.147786631005465e-06, + "loss": 0.036, + "step": 50180 + }, + { + "epoch": 1.5491697018334465, + "grad_norm": 0.3934556841850281, + "learning_rate": 7.1335941672803775e-06, + "loss": 0.035, + "step": 50190 + }, + { + "epoch": 1.549478362861905, + "grad_norm": 0.25327199697494507, + "learning_rate": 7.1194147252375384e-06, + "loss": 0.0365, + "step": 50200 + }, + { + "epoch": 1.5497870238903637, + "grad_norm": 0.36506661772727966, + "learning_rate": 7.10524830918432e-06, + "loss": 0.0318, + "step": 50210 + }, + { + "epoch": 1.550095684918822, + "grad_norm": 0.2517416775226593, + "learning_rate": 7.091094923424097e-06, + "loss": 0.0331, + "step": 50220 + }, + { + "epoch": 1.5504043459472807, + "grad_norm": 0.36689597368240356, + "learning_rate": 7.0769545722562894e-06, + "loss": 0.0346, + "step": 50230 + }, + { + "epoch": 1.550713006975739, + "grad_norm": 0.26482921838760376, + "learning_rate": 7.0628272599763675e-06, + "loss": 0.0315, + "step": 50240 + }, + { + "epoch": 1.5510216680041977, + "grad_norm": 0.2582803964614868, + "learning_rate": 7.048712990875828e-06, + "loss": 0.0298, + "step": 50250 + }, + { + "epoch": 1.5513303290326563, + "grad_norm": 0.2954101264476776, + "learning_rate": 7.034611769242216e-06, + "loss": 0.0337, + "step": 50260 + }, + { + "epoch": 1.551638990061115, + "grad_norm": 0.30941128730773926, + "learning_rate": 7.02052359935913e-06, + "loss": 0.0367, + "step": 50270 + }, + { + "epoch": 1.5519476510895736, + "grad_norm": 0.28098341822624207, + "learning_rate": 7.006448485506145e-06, + "loss": 0.0344, + "step": 50280 + }, + { + "epoch": 1.552256312118032, + "grad_norm": 0.34002071619033813, + "learning_rate": 6.992386431958942e-06, + "loss": 0.0321, + "step": 50290 + }, + { + "epoch": 1.5525649731464906, + "grad_norm": 0.27725473046302795, + "learning_rate": 6.978337442989197e-06, + "loss": 0.0335, + "step": 50300 + }, + { + "epoch": 1.552873634174949, + "grad_norm": 0.28845450282096863, + "learning_rate": 6.964301522864608e-06, + "loss": 0.0316, + "step": 50310 + }, + { + "epoch": 1.5531822952034076, + "grad_norm": 0.6370683312416077, + "learning_rate": 6.950278675848926e-06, + "loss": 0.0336, + "step": 50320 + }, + { + "epoch": 1.5534909562318662, + "grad_norm": 0.44364404678344727, + "learning_rate": 6.9362689062019145e-06, + "loss": 0.0336, + "step": 50330 + }, + { + "epoch": 1.5537996172603248, + "grad_norm": 0.38285112380981445, + "learning_rate": 6.922272218179393e-06, + "loss": 0.0344, + "step": 50340 + }, + { + "epoch": 1.5541082782887834, + "grad_norm": 0.39487341046333313, + "learning_rate": 6.908288616033148e-06, + "loss": 0.0316, + "step": 50350 + }, + { + "epoch": 1.5544169393172418, + "grad_norm": 0.34201911091804504, + "learning_rate": 6.894318104011077e-06, + "loss": 0.0305, + "step": 50360 + }, + { + "epoch": 1.5547256003457004, + "grad_norm": 0.6004717946052551, + "learning_rate": 6.880360686357007e-06, + "loss": 0.0356, + "step": 50370 + }, + { + "epoch": 1.5550342613741588, + "grad_norm": 0.3121965527534485, + "learning_rate": 6.8664163673108575e-06, + "loss": 0.0333, + "step": 50380 + }, + { + "epoch": 1.5553429224026174, + "grad_norm": 0.4331643283367157, + "learning_rate": 6.85248515110854e-06, + "loss": 0.0333, + "step": 50390 + }, + { + "epoch": 1.555651583431076, + "grad_norm": 0.2376362532377243, + "learning_rate": 6.838567041981992e-06, + "loss": 0.0334, + "step": 50400 + }, + { + "epoch": 1.5559602444595346, + "grad_norm": 0.24821412563323975, + "learning_rate": 6.8246620441591634e-06, + "loss": 0.0364, + "step": 50410 + }, + { + "epoch": 1.5562689054879932, + "grad_norm": 0.2935490310192108, + "learning_rate": 6.8107701618640275e-06, + "loss": 0.0353, + "step": 50420 + }, + { + "epoch": 1.5565775665164516, + "grad_norm": 0.277774453163147, + "learning_rate": 6.796891399316557e-06, + "loss": 0.0332, + "step": 50430 + }, + { + "epoch": 1.55688622754491, + "grad_norm": 0.27394992113113403, + "learning_rate": 6.7830257607327804e-06, + "loss": 0.0328, + "step": 50440 + }, + { + "epoch": 1.5571948885733686, + "grad_norm": 0.6801097989082336, + "learning_rate": 6.7691732503247e-06, + "loss": 0.0362, + "step": 50450 + }, + { + "epoch": 1.5575035496018272, + "grad_norm": 0.2770407199859619, + "learning_rate": 6.755333872300346e-06, + "loss": 0.036, + "step": 50460 + }, + { + "epoch": 1.5578122106302859, + "grad_norm": 0.5122328400611877, + "learning_rate": 6.741507630863747e-06, + "loss": 0.0349, + "step": 50470 + }, + { + "epoch": 1.5581208716587445, + "grad_norm": 0.3160146474838257, + "learning_rate": 6.727694530214945e-06, + "loss": 0.0351, + "step": 50480 + }, + { + "epoch": 1.558429532687203, + "grad_norm": 0.6230335831642151, + "learning_rate": 6.713894574550028e-06, + "loss": 0.035, + "step": 50490 + }, + { + "epoch": 1.5587381937156615, + "grad_norm": 0.28577283024787903, + "learning_rate": 6.700107768061015e-06, + "loss": 0.0327, + "step": 50500 + }, + { + "epoch": 1.5590468547441199, + "grad_norm": 0.47048744559288025, + "learning_rate": 6.686334114936016e-06, + "loss": 0.0359, + "step": 50510 + }, + { + "epoch": 1.5593555157725785, + "grad_norm": 0.2958971858024597, + "learning_rate": 6.672573619359063e-06, + "loss": 0.0337, + "step": 50520 + }, + { + "epoch": 1.559664176801037, + "grad_norm": 0.35583293437957764, + "learning_rate": 6.658826285510256e-06, + "loss": 0.0323, + "step": 50530 + }, + { + "epoch": 1.5599728378294957, + "grad_norm": 0.40513813495635986, + "learning_rate": 6.645092117565666e-06, + "loss": 0.033, + "step": 50540 + }, + { + "epoch": 1.5602814988579543, + "grad_norm": 0.3260266184806824, + "learning_rate": 6.631371119697371e-06, + "loss": 0.031, + "step": 50550 + }, + { + "epoch": 1.5605901598864127, + "grad_norm": 0.3800491988658905, + "learning_rate": 6.6176632960734505e-06, + "loss": 0.0371, + "step": 50560 + }, + { + "epoch": 1.5608988209148713, + "grad_norm": 0.37420666217803955, + "learning_rate": 6.603968650857978e-06, + "loss": 0.0333, + "step": 50570 + }, + { + "epoch": 1.5612074819433297, + "grad_norm": 0.2514200508594513, + "learning_rate": 6.5902871882110085e-06, + "loss": 0.0319, + "step": 50580 + }, + { + "epoch": 1.5615161429717883, + "grad_norm": 0.3402833938598633, + "learning_rate": 6.576618912288635e-06, + "loss": 0.0351, + "step": 50590 + }, + { + "epoch": 1.561824804000247, + "grad_norm": 0.3677261471748352, + "learning_rate": 6.562963827242913e-06, + "loss": 0.0359, + "step": 50600 + }, + { + "epoch": 1.5621334650287055, + "grad_norm": 0.2586193382740021, + "learning_rate": 6.549321937221886e-06, + "loss": 0.0372, + "step": 50610 + }, + { + "epoch": 1.5624421260571641, + "grad_norm": 0.24155041575431824, + "learning_rate": 6.5356932463696064e-06, + "loss": 0.0339, + "step": 50620 + }, + { + "epoch": 1.5627507870856225, + "grad_norm": 0.31930720806121826, + "learning_rate": 6.522077758826101e-06, + "loss": 0.0374, + "step": 50630 + }, + { + "epoch": 1.5630594481140812, + "grad_norm": 0.29833459854125977, + "learning_rate": 6.5084754787274275e-06, + "loss": 0.034, + "step": 50640 + }, + { + "epoch": 1.5633681091425395, + "grad_norm": 0.54181969165802, + "learning_rate": 6.494886410205553e-06, + "loss": 0.0356, + "step": 50650 + }, + { + "epoch": 1.5636767701709982, + "grad_norm": 0.3747994005680084, + "learning_rate": 6.481310557388521e-06, + "loss": 0.0297, + "step": 50660 + }, + { + "epoch": 1.5639854311994568, + "grad_norm": 0.3209502398967743, + "learning_rate": 6.46774792440028e-06, + "loss": 0.0365, + "step": 50670 + }, + { + "epoch": 1.5642940922279154, + "grad_norm": 0.3269694149494171, + "learning_rate": 6.4541985153608206e-06, + "loss": 0.0322, + "step": 50680 + }, + { + "epoch": 1.564602753256374, + "grad_norm": 0.29188090562820435, + "learning_rate": 6.4406623343861e-06, + "loss": 0.0332, + "step": 50690 + }, + { + "epoch": 1.5649114142848324, + "grad_norm": 0.25196099281311035, + "learning_rate": 6.427139385588038e-06, + "loss": 0.0339, + "step": 50700 + }, + { + "epoch": 1.565220075313291, + "grad_norm": 0.3528028726577759, + "learning_rate": 6.413629673074561e-06, + "loss": 0.0361, + "step": 50710 + }, + { + "epoch": 1.5655287363417494, + "grad_norm": 0.3027525544166565, + "learning_rate": 6.400133200949554e-06, + "loss": 0.0344, + "step": 50720 + }, + { + "epoch": 1.565837397370208, + "grad_norm": 0.39870336651802063, + "learning_rate": 6.386649973312897e-06, + "loss": 0.0322, + "step": 50730 + }, + { + "epoch": 1.5661460583986666, + "grad_norm": 0.34625911712646484, + "learning_rate": 6.37317999426042e-06, + "loss": 0.0339, + "step": 50740 + }, + { + "epoch": 1.5664547194271252, + "grad_norm": 0.32400187849998474, + "learning_rate": 6.359723267883977e-06, + "loss": 0.0331, + "step": 50750 + }, + { + "epoch": 1.5667633804555838, + "grad_norm": 0.26501989364624023, + "learning_rate": 6.346279798271343e-06, + "loss": 0.0312, + "step": 50760 + }, + { + "epoch": 1.5670720414840422, + "grad_norm": 0.39275112748146057, + "learning_rate": 6.332849589506301e-06, + "loss": 0.0369, + "step": 50770 + }, + { + "epoch": 1.5673807025125006, + "grad_norm": 0.47817087173461914, + "learning_rate": 6.319432645668588e-06, + "loss": 0.0366, + "step": 50780 + }, + { + "epoch": 1.5676893635409592, + "grad_norm": 0.3792741298675537, + "learning_rate": 6.306028970833922e-06, + "loss": 0.0344, + "step": 50790 + }, + { + "epoch": 1.5679980245694178, + "grad_norm": 0.2363186478614807, + "learning_rate": 6.2926385690739665e-06, + "loss": 0.0381, + "step": 50800 + }, + { + "epoch": 1.5683066855978764, + "grad_norm": 0.3433522582054138, + "learning_rate": 6.279261444456413e-06, + "loss": 0.0307, + "step": 50810 + }, + { + "epoch": 1.568615346626335, + "grad_norm": 0.2932112514972687, + "learning_rate": 6.265897601044829e-06, + "loss": 0.0355, + "step": 50820 + }, + { + "epoch": 1.5689240076547935, + "grad_norm": 0.2651960849761963, + "learning_rate": 6.2525470428988434e-06, + "loss": 0.0345, + "step": 50830 + }, + { + "epoch": 1.569232668683252, + "grad_norm": 0.2881161570549011, + "learning_rate": 6.239209774073962e-06, + "loss": 0.0337, + "step": 50840 + }, + { + "epoch": 1.5695413297117105, + "grad_norm": 0.3496306538581848, + "learning_rate": 6.225885798621728e-06, + "loss": 0.0335, + "step": 50850 + }, + { + "epoch": 1.569849990740169, + "grad_norm": 0.27642443776130676, + "learning_rate": 6.2125751205895925e-06, + "loss": 0.037, + "step": 50860 + }, + { + "epoch": 1.5701586517686277, + "grad_norm": 0.31786710023880005, + "learning_rate": 6.199277744020998e-06, + "loss": 0.0322, + "step": 50870 + }, + { + "epoch": 1.5704673127970863, + "grad_norm": 0.34453922510147095, + "learning_rate": 6.185993672955337e-06, + "loss": 0.0369, + "step": 50880 + }, + { + "epoch": 1.570775973825545, + "grad_norm": 0.41985026001930237, + "learning_rate": 6.172722911427947e-06, + "loss": 0.0341, + "step": 50890 + }, + { + "epoch": 1.5710846348540033, + "grad_norm": 0.35301199555397034, + "learning_rate": 6.159465463470149e-06, + "loss": 0.033, + "step": 50900 + }, + { + "epoch": 1.571393295882462, + "grad_norm": 0.3322874903678894, + "learning_rate": 6.146221333109204e-06, + "loss": 0.035, + "step": 50910 + }, + { + "epoch": 1.5717019569109203, + "grad_norm": 0.3303142786026001, + "learning_rate": 6.132990524368326e-06, + "loss": 0.0343, + "step": 50920 + }, + { + "epoch": 1.572010617939379, + "grad_norm": 0.242386132478714, + "learning_rate": 6.119773041266685e-06, + "loss": 0.0348, + "step": 50930 + }, + { + "epoch": 1.5723192789678375, + "grad_norm": 0.3543412983417511, + "learning_rate": 6.106568887819402e-06, + "loss": 0.0333, + "step": 50940 + }, + { + "epoch": 1.5726279399962961, + "grad_norm": 0.3520645201206207, + "learning_rate": 6.093378068037548e-06, + "loss": 0.0332, + "step": 50950 + }, + { + "epoch": 1.5729366010247547, + "grad_norm": 0.3435012996196747, + "learning_rate": 6.080200585928164e-06, + "loss": 0.0339, + "step": 50960 + }, + { + "epoch": 1.5732452620532131, + "grad_norm": 0.37270674109458923, + "learning_rate": 6.06703644549419e-06, + "loss": 0.0309, + "step": 50970 + }, + { + "epoch": 1.5735539230816717, + "grad_norm": 0.46164360642433167, + "learning_rate": 6.053885650734576e-06, + "loss": 0.0368, + "step": 50980 + }, + { + "epoch": 1.5738625841101301, + "grad_norm": 0.3268439769744873, + "learning_rate": 6.040748205644153e-06, + "loss": 0.0362, + "step": 50990 + }, + { + "epoch": 1.5741712451385887, + "grad_norm": 0.2421346753835678, + "learning_rate": 6.0276241142137646e-06, + "loss": 0.0327, + "step": 51000 + }, + { + "epoch": 1.5744799061670474, + "grad_norm": 0.8131046295166016, + "learning_rate": 6.014513380430142e-06, + "loss": 0.0306, + "step": 51010 + }, + { + "epoch": 1.574788567195506, + "grad_norm": 0.25808048248291016, + "learning_rate": 6.001416008275984e-06, + "loss": 0.0291, + "step": 51020 + }, + { + "epoch": 1.5750972282239646, + "grad_norm": 0.43882232904434204, + "learning_rate": 5.988332001729929e-06, + "loss": 0.0351, + "step": 51030 + }, + { + "epoch": 1.575405889252423, + "grad_norm": 0.24433131515979767, + "learning_rate": 5.975261364766543e-06, + "loss": 0.0368, + "step": 51040 + }, + { + "epoch": 1.5757145502808816, + "grad_norm": 0.2689829170703888, + "learning_rate": 5.962204101356356e-06, + "loss": 0.0342, + "step": 51050 + }, + { + "epoch": 1.57602321130934, + "grad_norm": 0.2504563331604004, + "learning_rate": 5.94916021546581e-06, + "loss": 0.0347, + "step": 51060 + }, + { + "epoch": 1.5763318723377986, + "grad_norm": 0.2737184464931488, + "learning_rate": 5.936129711057298e-06, + "loss": 0.033, + "step": 51070 + }, + { + "epoch": 1.5766405333662572, + "grad_norm": 0.30077043175697327, + "learning_rate": 5.923112592089142e-06, + "loss": 0.0335, + "step": 51080 + }, + { + "epoch": 1.5769491943947158, + "grad_norm": 0.43861666321754456, + "learning_rate": 5.9101088625155954e-06, + "loss": 0.0325, + "step": 51090 + }, + { + "epoch": 1.5772578554231744, + "grad_norm": 0.37083014845848083, + "learning_rate": 5.897118526286843e-06, + "loss": 0.0364, + "step": 51100 + }, + { + "epoch": 1.5775665164516328, + "grad_norm": 0.46447986364364624, + "learning_rate": 5.884141587349035e-06, + "loss": 0.0358, + "step": 51110 + }, + { + "epoch": 1.5778751774800912, + "grad_norm": 0.5320841073989868, + "learning_rate": 5.871178049644177e-06, + "loss": 0.0323, + "step": 51120 + }, + { + "epoch": 1.5781838385085498, + "grad_norm": 0.49052998423576355, + "learning_rate": 5.858227917110293e-06, + "loss": 0.0357, + "step": 51130 + }, + { + "epoch": 1.5784924995370084, + "grad_norm": 0.29189759492874146, + "learning_rate": 5.845291193681252e-06, + "loss": 0.0331, + "step": 51140 + }, + { + "epoch": 1.578801160565467, + "grad_norm": 0.500299334526062, + "learning_rate": 5.832367883286921e-06, + "loss": 0.0381, + "step": 51150 + }, + { + "epoch": 1.5791098215939257, + "grad_norm": 0.27883848547935486, + "learning_rate": 5.81945798985305e-06, + "loss": 0.0347, + "step": 51160 + }, + { + "epoch": 1.579418482622384, + "grad_norm": 0.462775856256485, + "learning_rate": 5.806561517301306e-06, + "loss": 0.0366, + "step": 51170 + }, + { + "epoch": 1.5797271436508427, + "grad_norm": 0.21865184605121613, + "learning_rate": 5.793678469549335e-06, + "loss": 0.0359, + "step": 51180 + }, + { + "epoch": 1.580035804679301, + "grad_norm": 0.27195248007774353, + "learning_rate": 5.780808850510627e-06, + "loss": 0.0352, + "step": 51190 + }, + { + "epoch": 1.5803444657077597, + "grad_norm": 0.32442373037338257, + "learning_rate": 5.767952664094673e-06, + "loss": 0.0319, + "step": 51200 + }, + { + "epoch": 1.5806531267362183, + "grad_norm": 0.29581418633461, + "learning_rate": 5.755109914206791e-06, + "loss": 0.033, + "step": 51210 + }, + { + "epoch": 1.5809617877646769, + "grad_norm": 0.3457561135292053, + "learning_rate": 5.7422806047483125e-06, + "loss": 0.035, + "step": 51220 + }, + { + "epoch": 1.5812704487931355, + "grad_norm": 0.2655464708805084, + "learning_rate": 5.72946473961643e-06, + "loss": 0.0345, + "step": 51230 + }, + { + "epoch": 1.5815791098215939, + "grad_norm": 0.38112911581993103, + "learning_rate": 5.716662322704264e-06, + "loss": 0.0355, + "step": 51240 + }, + { + "epoch": 1.5818877708500525, + "grad_norm": 0.3476918637752533, + "learning_rate": 5.703873357900852e-06, + "loss": 0.0342, + "step": 51250 + }, + { + "epoch": 1.582196431878511, + "grad_norm": 0.44369179010391235, + "learning_rate": 5.691097849091143e-06, + "loss": 0.0357, + "step": 51260 + }, + { + "epoch": 1.5825050929069695, + "grad_norm": 0.31224119663238525, + "learning_rate": 5.678335800155982e-06, + "loss": 0.0283, + "step": 51270 + }, + { + "epoch": 1.5828137539354281, + "grad_norm": 0.3015529215335846, + "learning_rate": 5.665587214972174e-06, + "loss": 0.0364, + "step": 51280 + }, + { + "epoch": 1.5831224149638867, + "grad_norm": 0.2728032171726227, + "learning_rate": 5.652852097412386e-06, + "loss": 0.0325, + "step": 51290 + }, + { + "epoch": 1.5834310759923453, + "grad_norm": 0.2831486761569977, + "learning_rate": 5.640130451345216e-06, + "loss": 0.0353, + "step": 51300 + }, + { + "epoch": 1.5837397370208037, + "grad_norm": 0.4344947040081024, + "learning_rate": 5.627422280635159e-06, + "loss": 0.032, + "step": 51310 + }, + { + "epoch": 1.5840483980492623, + "grad_norm": 0.3523130416870117, + "learning_rate": 5.61472758914261e-06, + "loss": 0.0322, + "step": 51320 + }, + { + "epoch": 1.5843570590777207, + "grad_norm": 0.3086232542991638, + "learning_rate": 5.602046380723918e-06, + "loss": 0.0324, + "step": 51330 + }, + { + "epoch": 1.5846657201061793, + "grad_norm": 0.4092749357223511, + "learning_rate": 5.5893786592312535e-06, + "loss": 0.0342, + "step": 51340 + }, + { + "epoch": 1.584974381134638, + "grad_norm": 0.48362600803375244, + "learning_rate": 5.576724428512775e-06, + "loss": 0.0348, + "step": 51350 + }, + { + "epoch": 1.5852830421630966, + "grad_norm": 0.2756921947002411, + "learning_rate": 5.56408369241247e-06, + "loss": 0.0313, + "step": 51360 + }, + { + "epoch": 1.5855917031915552, + "grad_norm": 0.5454574823379517, + "learning_rate": 5.5514564547702875e-06, + "loss": 0.0327, + "step": 51370 + }, + { + "epoch": 1.5859003642200136, + "grad_norm": 0.32816281914711, + "learning_rate": 5.538842719422038e-06, + "loss": 0.0359, + "step": 51380 + }, + { + "epoch": 1.5862090252484722, + "grad_norm": 0.508100152015686, + "learning_rate": 5.52624249019944e-06, + "loss": 0.0336, + "step": 51390 + }, + { + "epoch": 1.5865176862769306, + "grad_norm": 0.24358010292053223, + "learning_rate": 5.51365577093011e-06, + "loss": 0.0313, + "step": 51400 + }, + { + "epoch": 1.5868263473053892, + "grad_norm": 0.3058778941631317, + "learning_rate": 5.501082565437565e-06, + "loss": 0.0315, + "step": 51410 + }, + { + "epoch": 1.5871350083338478, + "grad_norm": 0.3818202614784241, + "learning_rate": 5.488522877541202e-06, + "loss": 0.0321, + "step": 51420 + }, + { + "epoch": 1.5874436693623064, + "grad_norm": 0.28176262974739075, + "learning_rate": 5.475976711056341e-06, + "loss": 0.0359, + "step": 51430 + }, + { + "epoch": 1.587752330390765, + "grad_norm": 0.3030446171760559, + "learning_rate": 5.463444069794166e-06, + "loss": 0.0324, + "step": 51440 + }, + { + "epoch": 1.5880609914192234, + "grad_norm": 0.2698506712913513, + "learning_rate": 5.4509249575617594e-06, + "loss": 0.0356, + "step": 51450 + }, + { + "epoch": 1.5883696524476818, + "grad_norm": 0.39336398243904114, + "learning_rate": 5.438419378162107e-06, + "loss": 0.0336, + "step": 51460 + }, + { + "epoch": 1.5886783134761404, + "grad_norm": 0.3995372951030731, + "learning_rate": 5.425927335394054e-06, + "loss": 0.0354, + "step": 51470 + }, + { + "epoch": 1.588986974504599, + "grad_norm": 0.43059587478637695, + "learning_rate": 5.413448833052387e-06, + "loss": 0.0341, + "step": 51480 + }, + { + "epoch": 1.5892956355330576, + "grad_norm": 0.28339239954948425, + "learning_rate": 5.400983874927701e-06, + "loss": 0.0351, + "step": 51490 + }, + { + "epoch": 1.5896042965615163, + "grad_norm": 0.2618892192840576, + "learning_rate": 5.388532464806567e-06, + "loss": 0.0363, + "step": 51500 + }, + { + "epoch": 1.5899129575899746, + "grad_norm": 0.3059471845626831, + "learning_rate": 5.3760946064713546e-06, + "loss": 0.0352, + "step": 51510 + }, + { + "epoch": 1.5902216186184333, + "grad_norm": 0.2989000678062439, + "learning_rate": 5.363670303700386e-06, + "loss": 0.0336, + "step": 51520 + }, + { + "epoch": 1.5905302796468916, + "grad_norm": 0.2875070869922638, + "learning_rate": 5.351259560267824e-06, + "loss": 0.0299, + "step": 51530 + }, + { + "epoch": 1.5908389406753503, + "grad_norm": 0.24648787081241608, + "learning_rate": 5.338862379943721e-06, + "loss": 0.0363, + "step": 51540 + }, + { + "epoch": 1.5911476017038089, + "grad_norm": 0.34462159872055054, + "learning_rate": 5.326478766494025e-06, + "loss": 0.0331, + "step": 51550 + }, + { + "epoch": 1.5914562627322675, + "grad_norm": 0.2940724194049835, + "learning_rate": 5.3141087236805385e-06, + "loss": 0.0346, + "step": 51560 + }, + { + "epoch": 1.591764923760726, + "grad_norm": 0.30533841252326965, + "learning_rate": 5.3017522552609615e-06, + "loss": 0.0346, + "step": 51570 + }, + { + "epoch": 1.5920735847891845, + "grad_norm": 0.266231894493103, + "learning_rate": 5.289409364988851e-06, + "loss": 0.0352, + "step": 51580 + }, + { + "epoch": 1.592382245817643, + "grad_norm": 0.5450723171234131, + "learning_rate": 5.277080056613671e-06, + "loss": 0.0356, + "step": 51590 + }, + { + "epoch": 1.5926909068461015, + "grad_norm": 0.26515859365463257, + "learning_rate": 5.264764333880729e-06, + "loss": 0.0336, + "step": 51600 + }, + { + "epoch": 1.59299956787456, + "grad_norm": 0.31211158633232117, + "learning_rate": 5.252462200531216e-06, + "loss": 0.0337, + "step": 51610 + }, + { + "epoch": 1.5933082289030187, + "grad_norm": 0.34677743911743164, + "learning_rate": 5.240173660302194e-06, + "loss": 0.036, + "step": 51620 + }, + { + "epoch": 1.5936168899314773, + "grad_norm": 0.3072843551635742, + "learning_rate": 5.2278987169266044e-06, + "loss": 0.0404, + "step": 51630 + }, + { + "epoch": 1.593925550959936, + "grad_norm": 0.316070556640625, + "learning_rate": 5.215637374133231e-06, + "loss": 0.0369, + "step": 51640 + }, + { + "epoch": 1.5942342119883943, + "grad_norm": 0.27735498547554016, + "learning_rate": 5.203389635646782e-06, + "loss": 0.0337, + "step": 51650 + }, + { + "epoch": 1.594542873016853, + "grad_norm": 0.3724878430366516, + "learning_rate": 5.191155505187756e-06, + "loss": 0.034, + "step": 51660 + }, + { + "epoch": 1.5948515340453113, + "grad_norm": 0.3591773808002472, + "learning_rate": 5.1789349864726e-06, + "loss": 0.0338, + "step": 51670 + }, + { + "epoch": 1.59516019507377, + "grad_norm": 0.3352603614330292, + "learning_rate": 5.16672808321354e-06, + "loss": 0.0354, + "step": 51680 + }, + { + "epoch": 1.5954688561022285, + "grad_norm": 0.3265039324760437, + "learning_rate": 5.154534799118749e-06, + "loss": 0.0341, + "step": 51690 + }, + { + "epoch": 1.5957775171306872, + "grad_norm": 0.3609347641468048, + "learning_rate": 5.142355137892207e-06, + "loss": 0.0317, + "step": 51700 + }, + { + "epoch": 1.5960861781591458, + "grad_norm": 0.31130629777908325, + "learning_rate": 5.130189103233779e-06, + "loss": 0.0353, + "step": 51710 + }, + { + "epoch": 1.5963948391876042, + "grad_norm": 0.42380291223526, + "learning_rate": 5.118036698839179e-06, + "loss": 0.0341, + "step": 51720 + }, + { + "epoch": 1.5967035002160626, + "grad_norm": 0.2458626627922058, + "learning_rate": 5.105897928399983e-06, + "loss": 0.0315, + "step": 51730 + }, + { + "epoch": 1.5970121612445212, + "grad_norm": 0.34618091583251953, + "learning_rate": 5.0937727956036464e-06, + "loss": 0.0386, + "step": 51740 + }, + { + "epoch": 1.5973208222729798, + "grad_norm": 0.32273587584495544, + "learning_rate": 5.081661304133456e-06, + "loss": 0.0322, + "step": 51750 + }, + { + "epoch": 1.5976294833014384, + "grad_norm": 0.4384361207485199, + "learning_rate": 5.069563457668558e-06, + "loss": 0.0322, + "step": 51760 + }, + { + "epoch": 1.597938144329897, + "grad_norm": 0.3461598753929138, + "learning_rate": 5.0574792598839624e-06, + "loss": 0.0333, + "step": 51770 + }, + { + "epoch": 1.5982468053583556, + "grad_norm": 0.24290509521961212, + "learning_rate": 5.0454087144505276e-06, + "loss": 0.029, + "step": 51780 + }, + { + "epoch": 1.598555466386814, + "grad_norm": 0.255214661359787, + "learning_rate": 5.0333518250349655e-06, + "loss": 0.0373, + "step": 51790 + }, + { + "epoch": 1.5988641274152724, + "grad_norm": 0.2861758768558502, + "learning_rate": 5.021308595299856e-06, + "loss": 0.0265, + "step": 51800 + }, + { + "epoch": 1.599172788443731, + "grad_norm": 0.32273170351982117, + "learning_rate": 5.009279028903585e-06, + "loss": 0.0316, + "step": 51810 + }, + { + "epoch": 1.5994814494721896, + "grad_norm": 0.254273384809494, + "learning_rate": 4.997263129500452e-06, + "loss": 0.033, + "step": 51820 + }, + { + "epoch": 1.5997901105006482, + "grad_norm": 0.22171775996685028, + "learning_rate": 4.985260900740535e-06, + "loss": 0.0353, + "step": 51830 + }, + { + "epoch": 1.6000987715291068, + "grad_norm": 0.33758267760276794, + "learning_rate": 4.973272346269814e-06, + "loss": 0.0355, + "step": 51840 + }, + { + "epoch": 1.6004074325575652, + "grad_norm": 0.3408306837081909, + "learning_rate": 4.961297469730097e-06, + "loss": 0.0342, + "step": 51850 + }, + { + "epoch": 1.6007160935860238, + "grad_norm": 0.24088646471500397, + "learning_rate": 4.949336274759031e-06, + "loss": 0.033, + "step": 51860 + }, + { + "epoch": 1.6010247546144822, + "grad_norm": 0.38550662994384766, + "learning_rate": 4.9373887649901144e-06, + "loss": 0.0353, + "step": 51870 + }, + { + "epoch": 1.6013334156429408, + "grad_norm": 0.3459989130496979, + "learning_rate": 4.925454944052666e-06, + "loss": 0.0318, + "step": 51880 + }, + { + "epoch": 1.6016420766713995, + "grad_norm": 0.28389236330986023, + "learning_rate": 4.913534815571891e-06, + "loss": 0.0291, + "step": 51890 + }, + { + "epoch": 1.601950737699858, + "grad_norm": 0.43864578008651733, + "learning_rate": 4.901628383168805e-06, + "loss": 0.0357, + "step": 51900 + }, + { + "epoch": 1.6022593987283167, + "grad_norm": 0.406796932220459, + "learning_rate": 4.8897356504602585e-06, + "loss": 0.0318, + "step": 51910 + }, + { + "epoch": 1.602568059756775, + "grad_norm": 0.3750194013118744, + "learning_rate": 4.877856621058957e-06, + "loss": 0.0322, + "step": 51920 + }, + { + "epoch": 1.6028767207852337, + "grad_norm": 0.47690191864967346, + "learning_rate": 4.86599129857343e-06, + "loss": 0.0315, + "step": 51930 + }, + { + "epoch": 1.603185381813692, + "grad_norm": 0.5727701783180237, + "learning_rate": 4.85413968660805e-06, + "loss": 0.0377, + "step": 51940 + }, + { + "epoch": 1.6034940428421507, + "grad_norm": 0.28301483392715454, + "learning_rate": 4.842301788763031e-06, + "loss": 0.0329, + "step": 51950 + }, + { + "epoch": 1.6038027038706093, + "grad_norm": 0.5673009157180786, + "learning_rate": 4.830477608634393e-06, + "loss": 0.0373, + "step": 51960 + }, + { + "epoch": 1.604111364899068, + "grad_norm": 0.3147479295730591, + "learning_rate": 4.818667149814049e-06, + "loss": 0.0327, + "step": 51970 + }, + { + "epoch": 1.6044200259275265, + "grad_norm": 0.5191996693611145, + "learning_rate": 4.806870415889664e-06, + "loss": 0.0351, + "step": 51980 + }, + { + "epoch": 1.604728686955985, + "grad_norm": 0.3780774772167206, + "learning_rate": 4.795087410444798e-06, + "loss": 0.0314, + "step": 51990 + }, + { + "epoch": 1.6050373479844435, + "grad_norm": 0.3574879467487335, + "learning_rate": 4.783318137058807e-06, + "loss": 0.0381, + "step": 52000 + }, + { + "epoch": 1.605346009012902, + "grad_norm": 0.2811161279678345, + "learning_rate": 4.771562599306895e-06, + "loss": 0.0307, + "step": 52010 + }, + { + "epoch": 1.6056546700413605, + "grad_norm": 0.3257031738758087, + "learning_rate": 4.759820800760073e-06, + "loss": 0.0307, + "step": 52020 + }, + { + "epoch": 1.6059633310698191, + "grad_norm": 0.39002349972724915, + "learning_rate": 4.7480927449851834e-06, + "loss": 0.0316, + "step": 52030 + }, + { + "epoch": 1.6062719920982778, + "grad_norm": 0.3562881648540497, + "learning_rate": 4.73637843554493e-06, + "loss": 0.0385, + "step": 52040 + }, + { + "epoch": 1.6065806531267364, + "grad_norm": 0.3289455771446228, + "learning_rate": 4.724677875997774e-06, + "loss": 0.0341, + "step": 52050 + }, + { + "epoch": 1.6068893141551948, + "grad_norm": 0.3971644937992096, + "learning_rate": 4.712991069898065e-06, + "loss": 0.035, + "step": 52060 + }, + { + "epoch": 1.6071979751836531, + "grad_norm": 0.34636178612709045, + "learning_rate": 4.7013180207959305e-06, + "loss": 0.0331, + "step": 52070 + }, + { + "epoch": 1.6075066362121118, + "grad_norm": 0.45862671732902527, + "learning_rate": 4.6896587322373395e-06, + "loss": 0.0288, + "step": 52080 + }, + { + "epoch": 1.6078152972405704, + "grad_norm": 0.32577162981033325, + "learning_rate": 4.678013207764081e-06, + "loss": 0.0356, + "step": 52090 + }, + { + "epoch": 1.608123958269029, + "grad_norm": 0.32700106501579285, + "learning_rate": 4.666381450913748e-06, + "loss": 0.0286, + "step": 52100 + }, + { + "epoch": 1.6084326192974876, + "grad_norm": 0.34897562861442566, + "learning_rate": 4.654763465219752e-06, + "loss": 0.0354, + "step": 52110 + }, + { + "epoch": 1.6087412803259462, + "grad_norm": 0.3424735367298126, + "learning_rate": 4.643159254211371e-06, + "loss": 0.0331, + "step": 52120 + }, + { + "epoch": 1.6090499413544046, + "grad_norm": 0.32042396068573, + "learning_rate": 4.631568821413606e-06, + "loss": 0.0379, + "step": 52130 + }, + { + "epoch": 1.609358602382863, + "grad_norm": 0.2772881090641022, + "learning_rate": 4.619992170347359e-06, + "loss": 0.0315, + "step": 52140 + }, + { + "epoch": 1.6096672634113216, + "grad_norm": 0.4262670874595642, + "learning_rate": 4.608429304529305e-06, + "loss": 0.0326, + "step": 52150 + }, + { + "epoch": 1.6099759244397802, + "grad_norm": 0.4351997971534729, + "learning_rate": 4.596880227471928e-06, + "loss": 0.0349, + "step": 52160 + }, + { + "epoch": 1.6102845854682388, + "grad_norm": 0.3163037896156311, + "learning_rate": 4.585344942683539e-06, + "loss": 0.0319, + "step": 52170 + }, + { + "epoch": 1.6105932464966974, + "grad_norm": 0.2891424596309662, + "learning_rate": 4.573823453668241e-06, + "loss": 0.0318, + "step": 52180 + }, + { + "epoch": 1.6109019075251558, + "grad_norm": 0.8205890655517578, + "learning_rate": 4.562315763925995e-06, + "loss": 0.0346, + "step": 52190 + }, + { + "epoch": 1.6112105685536144, + "grad_norm": 0.3571113646030426, + "learning_rate": 4.5508218769524825e-06, + "loss": 0.0323, + "step": 52200 + }, + { + "epoch": 1.6115192295820728, + "grad_norm": 0.3122173249721527, + "learning_rate": 4.539341796239277e-06, + "loss": 0.0334, + "step": 52210 + }, + { + "epoch": 1.6118278906105314, + "grad_norm": 0.30954277515411377, + "learning_rate": 4.527875525273717e-06, + "loss": 0.0363, + "step": 52220 + }, + { + "epoch": 1.61213655163899, + "grad_norm": 0.3268319368362427, + "learning_rate": 4.51642306753895e-06, + "loss": 0.0357, + "step": 52230 + }, + { + "epoch": 1.6124452126674487, + "grad_norm": 0.33763447403907776, + "learning_rate": 4.5049844265139306e-06, + "loss": 0.0324, + "step": 52240 + }, + { + "epoch": 1.6127538736959073, + "grad_norm": 0.25445711612701416, + "learning_rate": 4.4935596056734144e-06, + "loss": 0.0332, + "step": 52250 + }, + { + "epoch": 1.6130625347243657, + "grad_norm": 0.34267428517341614, + "learning_rate": 4.482148608487957e-06, + "loss": 0.0328, + "step": 52260 + }, + { + "epoch": 1.6133711957528243, + "grad_norm": 0.3170998692512512, + "learning_rate": 4.4707514384239365e-06, + "loss": 0.0341, + "step": 52270 + }, + { + "epoch": 1.6136798567812827, + "grad_norm": 0.2659134864807129, + "learning_rate": 4.459368098943484e-06, + "loss": 0.0319, + "step": 52280 + }, + { + "epoch": 1.6139885178097413, + "grad_norm": 0.29060792922973633, + "learning_rate": 4.447998593504582e-06, + "loss": 0.0343, + "step": 52290 + }, + { + "epoch": 1.6142971788382, + "grad_norm": 0.36743712425231934, + "learning_rate": 4.4366429255609744e-06, + "loss": 0.0344, + "step": 52300 + }, + { + "epoch": 1.6146058398666585, + "grad_norm": 0.28750288486480713, + "learning_rate": 4.425301098562212e-06, + "loss": 0.0326, + "step": 52310 + }, + { + "epoch": 1.6149145008951171, + "grad_norm": 0.31537723541259766, + "learning_rate": 4.413973115953651e-06, + "loss": 0.0343, + "step": 52320 + }, + { + "epoch": 1.6152231619235755, + "grad_norm": 0.3867699205875397, + "learning_rate": 4.402658981176416e-06, + "loss": 0.0359, + "step": 52330 + }, + { + "epoch": 1.6155318229520341, + "grad_norm": 0.405982106924057, + "learning_rate": 4.391358697667475e-06, + "loss": 0.0325, + "step": 52340 + }, + { + "epoch": 1.6158404839804925, + "grad_norm": 0.3645840585231781, + "learning_rate": 4.3800722688595195e-06, + "loss": 0.0314, + "step": 52350 + }, + { + "epoch": 1.6161491450089511, + "grad_norm": 0.30434179306030273, + "learning_rate": 4.368799698181097e-06, + "loss": 0.0341, + "step": 52360 + }, + { + "epoch": 1.6164578060374097, + "grad_norm": 0.29838836193084717, + "learning_rate": 4.357540989056486e-06, + "loss": 0.0352, + "step": 52370 + }, + { + "epoch": 1.6167664670658684, + "grad_norm": 0.27857705950737, + "learning_rate": 4.346296144905815e-06, + "loss": 0.037, + "step": 52380 + }, + { + "epoch": 1.617075128094327, + "grad_norm": 0.5055471062660217, + "learning_rate": 4.335065169144958e-06, + "loss": 0.0337, + "step": 52390 + }, + { + "epoch": 1.6173837891227854, + "grad_norm": 0.33143696188926697, + "learning_rate": 4.323848065185593e-06, + "loss": 0.0362, + "step": 52400 + }, + { + "epoch": 1.6176924501512437, + "grad_norm": 0.3562673032283783, + "learning_rate": 4.31264483643517e-06, + "loss": 0.0373, + "step": 52410 + }, + { + "epoch": 1.6180011111797024, + "grad_norm": 0.4214279353618622, + "learning_rate": 4.301455486296946e-06, + "loss": 0.0363, + "step": 52420 + }, + { + "epoch": 1.618309772208161, + "grad_norm": 0.2986413836479187, + "learning_rate": 4.290280018169935e-06, + "loss": 0.0342, + "step": 52430 + }, + { + "epoch": 1.6186184332366196, + "grad_norm": 0.33136582374572754, + "learning_rate": 4.27911843544897e-06, + "loss": 0.0338, + "step": 52440 + }, + { + "epoch": 1.6189270942650782, + "grad_norm": 0.4122694730758667, + "learning_rate": 4.2679707415246294e-06, + "loss": 0.0359, + "step": 52450 + }, + { + "epoch": 1.6192357552935366, + "grad_norm": 0.37879812717437744, + "learning_rate": 4.256836939783299e-06, + "loss": 0.033, + "step": 52460 + }, + { + "epoch": 1.6195444163219952, + "grad_norm": 0.356587678194046, + "learning_rate": 4.245717033607127e-06, + "loss": 0.0354, + "step": 52470 + }, + { + "epoch": 1.6198530773504536, + "grad_norm": 0.3429798185825348, + "learning_rate": 4.234611026374035e-06, + "loss": 0.0286, + "step": 52480 + }, + { + "epoch": 1.6201617383789122, + "grad_norm": 0.33738768100738525, + "learning_rate": 4.2235189214577694e-06, + "loss": 0.0356, + "step": 52490 + }, + { + "epoch": 1.6204703994073708, + "grad_norm": 0.4816349446773529, + "learning_rate": 4.212440722227779e-06, + "loss": 0.0349, + "step": 52500 + }, + { + "epoch": 1.6207790604358294, + "grad_norm": 0.41307497024536133, + "learning_rate": 4.201376432049364e-06, + "loss": 0.031, + "step": 52510 + }, + { + "epoch": 1.621087721464288, + "grad_norm": 0.24818524718284607, + "learning_rate": 4.1903260542835275e-06, + "loss": 0.0353, + "step": 52520 + }, + { + "epoch": 1.6213963824927464, + "grad_norm": 0.2900887131690979, + "learning_rate": 4.1792895922871114e-06, + "loss": 0.0326, + "step": 52530 + }, + { + "epoch": 1.621705043521205, + "grad_norm": 0.3507440686225891, + "learning_rate": 4.168267049412694e-06, + "loss": 0.0363, + "step": 52540 + }, + { + "epoch": 1.6220137045496634, + "grad_norm": 0.2673930823802948, + "learning_rate": 4.157258429008626e-06, + "loss": 0.0283, + "step": 52550 + }, + { + "epoch": 1.622322365578122, + "grad_norm": 0.25999850034713745, + "learning_rate": 4.146263734419043e-06, + "loss": 0.0313, + "step": 52560 + }, + { + "epoch": 1.6226310266065807, + "grad_norm": 0.37540194392204285, + "learning_rate": 4.135282968983839e-06, + "loss": 0.0396, + "step": 52570 + }, + { + "epoch": 1.6229396876350393, + "grad_norm": 0.3177950382232666, + "learning_rate": 4.124316136038675e-06, + "loss": 0.0328, + "step": 52580 + }, + { + "epoch": 1.6232483486634979, + "grad_norm": 0.430275022983551, + "learning_rate": 4.1133632389149965e-06, + "loss": 0.0392, + "step": 52590 + }, + { + "epoch": 1.6235570096919563, + "grad_norm": 0.35779255628585815, + "learning_rate": 4.102424280939998e-06, + "loss": 0.0329, + "step": 52600 + }, + { + "epoch": 1.6238656707204149, + "grad_norm": 0.33420902490615845, + "learning_rate": 4.091499265436649e-06, + "loss": 0.0311, + "step": 52610 + }, + { + "epoch": 1.6241743317488733, + "grad_norm": 0.31393444538116455, + "learning_rate": 4.080588195723684e-06, + "loss": 0.0361, + "step": 52620 + }, + { + "epoch": 1.6244829927773319, + "grad_norm": 0.23028558492660522, + "learning_rate": 4.069691075115578e-06, + "loss": 0.0353, + "step": 52630 + }, + { + "epoch": 1.6247916538057905, + "grad_norm": 0.23461481928825378, + "learning_rate": 4.0588079069226235e-06, + "loss": 0.0369, + "step": 52640 + }, + { + "epoch": 1.625100314834249, + "grad_norm": 0.2545364499092102, + "learning_rate": 4.0479386944508034e-06, + "loss": 0.0345, + "step": 52650 + }, + { + "epoch": 1.6254089758627077, + "grad_norm": 0.3129882514476776, + "learning_rate": 4.037083441001932e-06, + "loss": 0.038, + "step": 52660 + }, + { + "epoch": 1.625717636891166, + "grad_norm": 0.3119959533214569, + "learning_rate": 4.026242149873516e-06, + "loss": 0.0385, + "step": 52670 + }, + { + "epoch": 1.6260262979196247, + "grad_norm": 0.3694458305835724, + "learning_rate": 4.015414824358871e-06, + "loss": 0.0354, + "step": 52680 + }, + { + "epoch": 1.626334958948083, + "grad_norm": 0.35517311096191406, + "learning_rate": 4.004601467747054e-06, + "loss": 0.0343, + "step": 52690 + }, + { + "epoch": 1.6266436199765417, + "grad_norm": 0.30639538168907166, + "learning_rate": 3.993802083322873e-06, + "loss": 0.0331, + "step": 52700 + }, + { + "epoch": 1.6269522810050003, + "grad_norm": 0.3799438178539276, + "learning_rate": 3.9830166743668906e-06, + "loss": 0.0369, + "step": 52710 + }, + { + "epoch": 1.627260942033459, + "grad_norm": 0.44921159744262695, + "learning_rate": 3.9722452441554425e-06, + "loss": 0.0355, + "step": 52720 + }, + { + "epoch": 1.6275696030619176, + "grad_norm": 0.34430038928985596, + "learning_rate": 3.961487795960584e-06, + "loss": 0.0356, + "step": 52730 + }, + { + "epoch": 1.627878264090376, + "grad_norm": 0.32792046666145325, + "learning_rate": 3.95074433305016e-06, + "loss": 0.0318, + "step": 52740 + }, + { + "epoch": 1.6281869251188343, + "grad_norm": 0.3611304759979248, + "learning_rate": 3.940014858687752e-06, + "loss": 0.0348, + "step": 52750 + }, + { + "epoch": 1.628495586147293, + "grad_norm": 0.39409443736076355, + "learning_rate": 3.929299376132689e-06, + "loss": 0.0336, + "step": 52760 + }, + { + "epoch": 1.6288042471757516, + "grad_norm": 0.3291214406490326, + "learning_rate": 3.918597888640047e-06, + "loss": 0.0315, + "step": 52770 + }, + { + "epoch": 1.6291129082042102, + "grad_norm": 0.24573072791099548, + "learning_rate": 3.907910399460657e-06, + "loss": 0.035, + "step": 52780 + }, + { + "epoch": 1.6294215692326688, + "grad_norm": 0.2822451591491699, + "learning_rate": 3.8972369118410956e-06, + "loss": 0.0313, + "step": 52790 + }, + { + "epoch": 1.6297302302611272, + "grad_norm": 0.28250977396965027, + "learning_rate": 3.88657742902368e-06, + "loss": 0.0327, + "step": 52800 + }, + { + "epoch": 1.6300388912895858, + "grad_norm": 0.2866600453853607, + "learning_rate": 3.875931954246504e-06, + "loss": 0.0312, + "step": 52810 + }, + { + "epoch": 1.6303475523180442, + "grad_norm": 0.2650095224380493, + "learning_rate": 3.865300490743351e-06, + "loss": 0.0365, + "step": 52820 + }, + { + "epoch": 1.6306562133465028, + "grad_norm": 0.3217502236366272, + "learning_rate": 3.854683041743806e-06, + "loss": 0.0344, + "step": 52830 + }, + { + "epoch": 1.6309648743749614, + "grad_norm": 0.3531551659107208, + "learning_rate": 3.844079610473139e-06, + "loss": 0.0343, + "step": 52840 + }, + { + "epoch": 1.63127353540342, + "grad_norm": 0.7148820161819458, + "learning_rate": 3.833490200152423e-06, + "loss": 0.0341, + "step": 52850 + }, + { + "epoch": 1.6315821964318786, + "grad_norm": 0.28896763920783997, + "learning_rate": 3.822914813998424e-06, + "loss": 0.0329, + "step": 52860 + }, + { + "epoch": 1.631890857460337, + "grad_norm": 0.24091587960720062, + "learning_rate": 3.812353455223666e-06, + "loss": 0.0353, + "step": 52870 + }, + { + "epoch": 1.6321995184887956, + "grad_norm": 0.32536977529525757, + "learning_rate": 3.8018061270364225e-06, + "loss": 0.0333, + "step": 52880 + }, + { + "epoch": 1.632508179517254, + "grad_norm": 0.3517014980316162, + "learning_rate": 3.7912728326406688e-06, + "loss": 0.0329, + "step": 52890 + }, + { + "epoch": 1.6328168405457126, + "grad_norm": 0.2748081088066101, + "learning_rate": 3.7807535752361732e-06, + "loss": 0.0373, + "step": 52900 + }, + { + "epoch": 1.6331255015741712, + "grad_norm": 0.2797493636608124, + "learning_rate": 3.7702483580183855e-06, + "loss": 0.0311, + "step": 52910 + }, + { + "epoch": 1.6334341626026299, + "grad_norm": 0.4345960319042206, + "learning_rate": 3.759757184178525e-06, + "loss": 0.0309, + "step": 52920 + }, + { + "epoch": 1.6337428236310885, + "grad_norm": 0.2770526707172394, + "learning_rate": 3.7492800569035312e-06, + "loss": 0.0314, + "step": 52930 + }, + { + "epoch": 1.6340514846595469, + "grad_norm": 0.908765435218811, + "learning_rate": 3.7388169793760754e-06, + "loss": 0.0366, + "step": 52940 + }, + { + "epoch": 1.6343601456880055, + "grad_norm": 0.4564135670661926, + "learning_rate": 3.728367954774553e-06, + "loss": 0.0359, + "step": 52950 + }, + { + "epoch": 1.6346688067164639, + "grad_norm": 0.3026871085166931, + "learning_rate": 3.7179329862731317e-06, + "loss": 0.0357, + "step": 52960 + }, + { + "epoch": 1.6349774677449225, + "grad_norm": 0.2979544699192047, + "learning_rate": 3.707512077041647e-06, + "loss": 0.0353, + "step": 52970 + }, + { + "epoch": 1.635286128773381, + "grad_norm": 5.886374473571777, + "learning_rate": 3.6971052302457288e-06, + "loss": 0.0333, + "step": 52980 + }, + { + "epoch": 1.6355947898018397, + "grad_norm": 0.23849599063396454, + "learning_rate": 3.6867124490466698e-06, + "loss": 0.0314, + "step": 52990 + }, + { + "epoch": 1.6359034508302983, + "grad_norm": 0.296614408493042, + "learning_rate": 3.6763337366015393e-06, + "loss": 0.0331, + "step": 53000 + }, + { + "epoch": 1.6362121118587567, + "grad_norm": 0.38481324911117554, + "learning_rate": 3.665969096063121e-06, + "loss": 0.0315, + "step": 53010 + }, + { + "epoch": 1.6365207728872153, + "grad_norm": 0.3536874055862427, + "learning_rate": 3.6556185305799074e-06, + "loss": 0.0347, + "step": 53020 + }, + { + "epoch": 1.6368294339156737, + "grad_norm": 0.2629459500312805, + "learning_rate": 3.645282043296133e-06, + "loss": 0.0359, + "step": 53030 + }, + { + "epoch": 1.6371380949441323, + "grad_norm": 0.2537602484226227, + "learning_rate": 3.6349596373517427e-06, + "loss": 0.0385, + "step": 53040 + }, + { + "epoch": 1.637446755972591, + "grad_norm": 0.3727093040943146, + "learning_rate": 3.6246513158824215e-06, + "loss": 0.036, + "step": 53050 + }, + { + "epoch": 1.6377554170010495, + "grad_norm": 0.3808312714099884, + "learning_rate": 3.6143570820195593e-06, + "loss": 0.0362, + "step": 53060 + }, + { + "epoch": 1.6380640780295082, + "grad_norm": 0.36152157187461853, + "learning_rate": 3.6040769388902773e-06, + "loss": 0.0327, + "step": 53070 + }, + { + "epoch": 1.6383727390579665, + "grad_norm": 0.2710995674133301, + "learning_rate": 3.593810889617405e-06, + "loss": 0.0334, + "step": 53080 + }, + { + "epoch": 1.638681400086425, + "grad_norm": 0.3042760491371155, + "learning_rate": 3.5835589373194978e-06, + "loss": 0.0287, + "step": 53090 + }, + { + "epoch": 1.6389900611148835, + "grad_norm": 0.42022937536239624, + "learning_rate": 3.5733210851108257e-06, + "loss": 0.0349, + "step": 53100 + }, + { + "epoch": 1.6392987221433422, + "grad_norm": 0.2886767089366913, + "learning_rate": 3.5630973361014008e-06, + "loss": 0.0328, + "step": 53110 + }, + { + "epoch": 1.6396073831718008, + "grad_norm": 0.4077942371368408, + "learning_rate": 3.552887693396889e-06, + "loss": 0.0371, + "step": 53120 + }, + { + "epoch": 1.6399160442002594, + "grad_norm": 0.6965972781181335, + "learning_rate": 3.542692160098754e-06, + "loss": 0.0338, + "step": 53130 + }, + { + "epoch": 1.6402247052287178, + "grad_norm": 0.3033323287963867, + "learning_rate": 3.5325107393040846e-06, + "loss": 0.0332, + "step": 53140 + }, + { + "epoch": 1.6405333662571764, + "grad_norm": 0.27698588371276855, + "learning_rate": 3.522343434105757e-06, + "loss": 0.0289, + "step": 53150 + }, + { + "epoch": 1.6408420272856348, + "grad_norm": 0.31630903482437134, + "learning_rate": 3.512190247592323e-06, + "loss": 0.0301, + "step": 53160 + }, + { + "epoch": 1.6411506883140934, + "grad_norm": 0.2626039981842041, + "learning_rate": 3.502051182848054e-06, + "loss": 0.0315, + "step": 53170 + }, + { + "epoch": 1.641459349342552, + "grad_norm": 0.2988065779209137, + "learning_rate": 3.4919262429529308e-06, + "loss": 0.0333, + "step": 53180 + }, + { + "epoch": 1.6417680103710106, + "grad_norm": 0.2678147852420807, + "learning_rate": 3.4818154309826325e-06, + "loss": 0.0296, + "step": 53190 + }, + { + "epoch": 1.6420766713994692, + "grad_norm": 0.3075389564037323, + "learning_rate": 3.4717187500085734e-06, + "loss": 0.0362, + "step": 53200 + }, + { + "epoch": 1.6423853324279276, + "grad_norm": 0.29710304737091064, + "learning_rate": 3.46163620309784e-06, + "loss": 0.0323, + "step": 53210 + }, + { + "epoch": 1.6426939934563862, + "grad_norm": 0.4137555658817291, + "learning_rate": 3.4515677933132595e-06, + "loss": 0.0326, + "step": 53220 + }, + { + "epoch": 1.6430026544848446, + "grad_norm": 0.2545850872993469, + "learning_rate": 3.4415135237133466e-06, + "loss": 0.0318, + "step": 53230 + }, + { + "epoch": 1.6433113155133032, + "grad_norm": 0.3071592450141907, + "learning_rate": 3.4314733973523196e-06, + "loss": 0.0364, + "step": 53240 + }, + { + "epoch": 1.6436199765417618, + "grad_norm": 0.2493392825126648, + "learning_rate": 3.4214474172800993e-06, + "loss": 0.0356, + "step": 53250 + }, + { + "epoch": 1.6439286375702205, + "grad_norm": 0.29622313380241394, + "learning_rate": 3.411435586542322e-06, + "loss": 0.0341, + "step": 53260 + }, + { + "epoch": 1.644237298598679, + "grad_norm": 0.3544348478317261, + "learning_rate": 3.4014379081802995e-06, + "loss": 0.0364, + "step": 53270 + }, + { + "epoch": 1.6445459596271375, + "grad_norm": 0.27684131264686584, + "learning_rate": 3.391454385231102e-06, + "loss": 0.0317, + "step": 53280 + }, + { + "epoch": 1.644854620655596, + "grad_norm": 0.37532275915145874, + "learning_rate": 3.3814850207274095e-06, + "loss": 0.0344, + "step": 53290 + }, + { + "epoch": 1.6451632816840545, + "grad_norm": 0.2138802707195282, + "learning_rate": 3.371529817697694e-06, + "loss": 0.0332, + "step": 53300 + }, + { + "epoch": 1.645471942712513, + "grad_norm": 0.46436452865600586, + "learning_rate": 3.3615887791660585e-06, + "loss": 0.036, + "step": 53310 + }, + { + "epoch": 1.6457806037409717, + "grad_norm": 0.3158811628818512, + "learning_rate": 3.3516619081523426e-06, + "loss": 0.0392, + "step": 53320 + }, + { + "epoch": 1.6460892647694303, + "grad_norm": 0.4348369538784027, + "learning_rate": 3.3417492076720567e-06, + "loss": 0.0321, + "step": 53330 + }, + { + "epoch": 1.646397925797889, + "grad_norm": 0.2975000739097595, + "learning_rate": 3.3318506807364147e-06, + "loss": 0.035, + "step": 53340 + }, + { + "epoch": 1.6467065868263473, + "grad_norm": 0.2760465443134308, + "learning_rate": 3.3219663303523553e-06, + "loss": 0.0352, + "step": 53350 + }, + { + "epoch": 1.6470152478548057, + "grad_norm": 0.2650717496871948, + "learning_rate": 3.3120961595224374e-06, + "loss": 0.036, + "step": 53360 + }, + { + "epoch": 1.6473239088832643, + "grad_norm": 0.2706119120121002, + "learning_rate": 3.302240171245002e-06, + "loss": 0.0344, + "step": 53370 + }, + { + "epoch": 1.647632569911723, + "grad_norm": 0.5066658854484558, + "learning_rate": 3.29239836851401e-06, + "loss": 0.0317, + "step": 53380 + }, + { + "epoch": 1.6479412309401815, + "grad_norm": 0.5481063723564148, + "learning_rate": 3.2825707543191588e-06, + "loss": 0.0333, + "step": 53390 + }, + { + "epoch": 1.6482498919686401, + "grad_norm": 0.34326085448265076, + "learning_rate": 3.272757331645804e-06, + "loss": 0.0343, + "step": 53400 + }, + { + "epoch": 1.6485585529970987, + "grad_norm": 0.28348660469055176, + "learning_rate": 3.2629581034750166e-06, + "loss": 0.0324, + "step": 53410 + }, + { + "epoch": 1.6488672140255571, + "grad_norm": 0.3828966021537781, + "learning_rate": 3.2531730727835218e-06, + "loss": 0.0347, + "step": 53420 + }, + { + "epoch": 1.6491758750540155, + "grad_norm": 0.32572320103645325, + "learning_rate": 3.2434022425437914e-06, + "loss": 0.0298, + "step": 53430 + }, + { + "epoch": 1.6494845360824741, + "grad_norm": 0.31915977597236633, + "learning_rate": 3.233645615723907e-06, + "loss": 0.0323, + "step": 53440 + }, + { + "epoch": 1.6497931971109328, + "grad_norm": 0.33035802841186523, + "learning_rate": 3.2239031952876918e-06, + "loss": 0.0311, + "step": 53450 + }, + { + "epoch": 1.6501018581393914, + "grad_norm": 0.29441243410110474, + "learning_rate": 3.21417498419464e-06, + "loss": 0.0335, + "step": 53460 + }, + { + "epoch": 1.65041051916785, + "grad_norm": 0.37509915232658386, + "learning_rate": 3.204460985399921e-06, + "loss": 0.0335, + "step": 53470 + }, + { + "epoch": 1.6507191801963084, + "grad_norm": 0.22756490111351013, + "learning_rate": 3.1947612018543903e-06, + "loss": 0.0324, + "step": 53480 + }, + { + "epoch": 1.651027841224767, + "grad_norm": 0.22407731413841248, + "learning_rate": 3.1850756365045753e-06, + "loss": 0.0285, + "step": 53490 + }, + { + "epoch": 1.6513365022532254, + "grad_norm": 0.3647134602069855, + "learning_rate": 3.175404292292722e-06, + "loss": 0.0342, + "step": 53500 + }, + { + "epoch": 1.651645163281684, + "grad_norm": 0.29919537901878357, + "learning_rate": 3.1657471721566965e-06, + "loss": 0.0341, + "step": 53510 + }, + { + "epoch": 1.6519538243101426, + "grad_norm": 0.29262474179267883, + "learning_rate": 3.1561042790300977e-06, + "loss": 0.038, + "step": 53520 + }, + { + "epoch": 1.6522624853386012, + "grad_norm": 0.47436678409576416, + "learning_rate": 3.1464756158421816e-06, + "loss": 0.0305, + "step": 53530 + }, + { + "epoch": 1.6525711463670598, + "grad_norm": 0.46008023619651794, + "learning_rate": 3.136861185517875e-06, + "loss": 0.0318, + "step": 53540 + }, + { + "epoch": 1.6528798073955182, + "grad_norm": 0.4005986452102661, + "learning_rate": 3.127260990977798e-06, + "loss": 0.0361, + "step": 53550 + }, + { + "epoch": 1.6531884684239768, + "grad_norm": 0.2868867516517639, + "learning_rate": 3.1176750351382235e-06, + "loss": 0.0376, + "step": 53560 + }, + { + "epoch": 1.6534971294524352, + "grad_norm": 0.41461583971977234, + "learning_rate": 3.1081033209111153e-06, + "loss": 0.0336, + "step": 53570 + }, + { + "epoch": 1.6538057904808938, + "grad_norm": 0.5283703207969666, + "learning_rate": 3.0985458512041155e-06, + "loss": 0.0337, + "step": 53580 + }, + { + "epoch": 1.6541144515093524, + "grad_norm": 0.3459003269672394, + "learning_rate": 3.089002628920512e-06, + "loss": 0.0342, + "step": 53590 + }, + { + "epoch": 1.654423112537811, + "grad_norm": 0.38875922560691833, + "learning_rate": 3.079473656959303e-06, + "loss": 0.0338, + "step": 53600 + }, + { + "epoch": 1.6547317735662697, + "grad_norm": 0.24435897171497345, + "learning_rate": 3.0699589382151393e-06, + "loss": 0.033, + "step": 53610 + }, + { + "epoch": 1.655040434594728, + "grad_norm": 0.3967825174331665, + "learning_rate": 3.060458475578326e-06, + "loss": 0.0361, + "step": 53620 + }, + { + "epoch": 1.6553490956231867, + "grad_norm": 0.4564705491065979, + "learning_rate": 3.0509722719348656e-06, + "loss": 0.0362, + "step": 53630 + }, + { + "epoch": 1.655657756651645, + "grad_norm": 0.4004043638706207, + "learning_rate": 3.041500330166408e-06, + "loss": 0.0312, + "step": 53640 + }, + { + "epoch": 1.6559664176801037, + "grad_norm": 0.3489733338356018, + "learning_rate": 3.032042653150291e-06, + "loss": 0.0313, + "step": 53650 + }, + { + "epoch": 1.6562750787085623, + "grad_norm": 0.4086248576641083, + "learning_rate": 3.0225992437594887e-06, + "loss": 0.0333, + "step": 53660 + }, + { + "epoch": 1.6565837397370209, + "grad_norm": 0.35132575035095215, + "learning_rate": 3.0131701048626783e-06, + "loss": 0.0346, + "step": 53670 + }, + { + "epoch": 1.6568924007654795, + "grad_norm": 0.3480113744735718, + "learning_rate": 3.003755239324163e-06, + "loss": 0.0392, + "step": 53680 + }, + { + "epoch": 1.657201061793938, + "grad_norm": 0.26192131638526917, + "learning_rate": 2.9943546500039553e-06, + "loss": 0.0342, + "step": 53690 + }, + { + "epoch": 1.6575097228223963, + "grad_norm": 0.34076204895973206, + "learning_rate": 2.9849683397576877e-06, + "loss": 0.0351, + "step": 53700 + }, + { + "epoch": 1.657818383850855, + "grad_norm": 0.33470097184181213, + "learning_rate": 2.975596311436679e-06, + "loss": 0.0362, + "step": 53710 + }, + { + "epoch": 1.6581270448793135, + "grad_norm": 0.36260876059532166, + "learning_rate": 2.966238567887902e-06, + "loss": 0.0338, + "step": 53720 + }, + { + "epoch": 1.6584357059077721, + "grad_norm": 0.3202480375766754, + "learning_rate": 2.9568951119539943e-06, + "loss": 0.0331, + "step": 53730 + }, + { + "epoch": 1.6587443669362307, + "grad_norm": 0.5321466326713562, + "learning_rate": 2.947565946473241e-06, + "loss": 0.0332, + "step": 53740 + }, + { + "epoch": 1.6590530279646893, + "grad_norm": 0.2294619232416153, + "learning_rate": 2.9382510742796188e-06, + "loss": 0.0299, + "step": 53750 + }, + { + "epoch": 1.6593616889931477, + "grad_norm": 0.3185577690601349, + "learning_rate": 2.9289504982027204e-06, + "loss": 0.0346, + "step": 53760 + }, + { + "epoch": 1.6596703500216061, + "grad_norm": 0.2949177920818329, + "learning_rate": 2.9196642210678248e-06, + "loss": 0.0358, + "step": 53770 + }, + { + "epoch": 1.6599790110500647, + "grad_norm": 0.3109104633331299, + "learning_rate": 2.910392245695853e-06, + "loss": 0.0317, + "step": 53780 + }, + { + "epoch": 1.6602876720785233, + "grad_norm": 0.32221099734306335, + "learning_rate": 2.9011345749033804e-06, + "loss": 0.0328, + "step": 53790 + }, + { + "epoch": 1.660596333106982, + "grad_norm": 0.2515459656715393, + "learning_rate": 2.8918912115026677e-06, + "loss": 0.0333, + "step": 53800 + }, + { + "epoch": 1.6609049941354406, + "grad_norm": 0.3881371021270752, + "learning_rate": 2.8826621583015743e-06, + "loss": 0.0353, + "step": 53810 + }, + { + "epoch": 1.661213655163899, + "grad_norm": 0.32144874334335327, + "learning_rate": 2.8734474181036643e-06, + "loss": 0.0307, + "step": 53820 + }, + { + "epoch": 1.6615223161923576, + "grad_norm": 0.3480967581272125, + "learning_rate": 2.864246993708114e-06, + "loss": 0.0324, + "step": 53830 + }, + { + "epoch": 1.661830977220816, + "grad_norm": 0.26661086082458496, + "learning_rate": 2.8550608879097884e-06, + "loss": 0.031, + "step": 53840 + }, + { + "epoch": 1.6621396382492746, + "grad_norm": 0.24576561152935028, + "learning_rate": 2.845889103499172e-06, + "loss": 0.0355, + "step": 53850 + }, + { + "epoch": 1.6624482992777332, + "grad_norm": 0.619060754776001, + "learning_rate": 2.8367316432624138e-06, + "loss": 0.031, + "step": 53860 + }, + { + "epoch": 1.6627569603061918, + "grad_norm": 0.3721485733985901, + "learning_rate": 2.8275885099813105e-06, + "loss": 0.0373, + "step": 53870 + }, + { + "epoch": 1.6630656213346504, + "grad_norm": 0.35913321375846863, + "learning_rate": 2.8184597064332963e-06, + "loss": 0.0355, + "step": 53880 + }, + { + "epoch": 1.6633742823631088, + "grad_norm": 0.2549053728580475, + "learning_rate": 2.809345235391464e-06, + "loss": 0.0341, + "step": 53890 + }, + { + "epoch": 1.6636829433915674, + "grad_norm": 0.32153522968292236, + "learning_rate": 2.80024509962456e-06, + "loss": 0.0356, + "step": 53900 + }, + { + "epoch": 1.6639916044200258, + "grad_norm": 0.43460917472839355, + "learning_rate": 2.7911593018969563e-06, + "loss": 0.0372, + "step": 53910 + }, + { + "epoch": 1.6643002654484844, + "grad_norm": 0.33850952982902527, + "learning_rate": 2.7820878449686838e-06, + "loss": 0.0368, + "step": 53920 + }, + { + "epoch": 1.664608926476943, + "grad_norm": 0.3162692189216614, + "learning_rate": 2.7730307315953995e-06, + "loss": 0.035, + "step": 53930 + }, + { + "epoch": 1.6649175875054016, + "grad_norm": 0.39138704538345337, + "learning_rate": 2.763987964528425e-06, + "loss": 0.0314, + "step": 53940 + }, + { + "epoch": 1.6652262485338603, + "grad_norm": 0.44808512926101685, + "learning_rate": 2.754959546514718e-06, + "loss": 0.0348, + "step": 53950 + }, + { + "epoch": 1.6655349095623186, + "grad_norm": 0.2986275255680084, + "learning_rate": 2.7459454802968576e-06, + "loss": 0.03, + "step": 53960 + }, + { + "epoch": 1.6658435705907773, + "grad_norm": 0.39772871136665344, + "learning_rate": 2.7369457686131028e-06, + "loss": 0.0405, + "step": 53970 + }, + { + "epoch": 1.6661522316192356, + "grad_norm": 0.3185187876224518, + "learning_rate": 2.7279604141973004e-06, + "loss": 0.0346, + "step": 53980 + }, + { + "epoch": 1.6664608926476943, + "grad_norm": 0.30142977833747864, + "learning_rate": 2.7189894197789946e-06, + "loss": 0.036, + "step": 53990 + }, + { + "epoch": 1.6667695536761529, + "grad_norm": 0.3019247353076935, + "learning_rate": 2.7100327880833055e-06, + "loss": 0.0315, + "step": 54000 + }, + { + "epoch": 1.6670782147046115, + "grad_norm": 0.8488101363182068, + "learning_rate": 2.70109052183104e-06, + "loss": 0.034, + "step": 54010 + }, + { + "epoch": 1.66738687573307, + "grad_norm": 0.22734646499156952, + "learning_rate": 2.69216262373862e-06, + "loss": 0.0304, + "step": 54020 + }, + { + "epoch": 1.6676955367615285, + "grad_norm": 0.3200460970401764, + "learning_rate": 2.6832490965181036e-06, + "loss": 0.0327, + "step": 54030 + }, + { + "epoch": 1.6680041977899869, + "grad_norm": 0.3277895152568817, + "learning_rate": 2.6743499428771857e-06, + "loss": 0.0308, + "step": 54040 + }, + { + "epoch": 1.6683128588184455, + "grad_norm": 0.5101187825202942, + "learning_rate": 2.6654651655191875e-06, + "loss": 0.0338, + "step": 54050 + }, + { + "epoch": 1.668621519846904, + "grad_norm": 0.2601391077041626, + "learning_rate": 2.6565947671430836e-06, + "loss": 0.0372, + "step": 54060 + }, + { + "epoch": 1.6689301808753627, + "grad_norm": 0.4148057997226715, + "learning_rate": 2.647738750443457e-06, + "loss": 0.0336, + "step": 54070 + }, + { + "epoch": 1.6692388419038213, + "grad_norm": 0.30460357666015625, + "learning_rate": 2.6388971181105393e-06, + "loss": 0.0337, + "step": 54080 + }, + { + "epoch": 1.66954750293228, + "grad_norm": 0.35598042607307434, + "learning_rate": 2.630069872830171e-06, + "loss": 0.0334, + "step": 54090 + }, + { + "epoch": 1.6698561639607383, + "grad_norm": 0.41696953773498535, + "learning_rate": 2.6212570172838514e-06, + "loss": 0.0359, + "step": 54100 + }, + { + "epoch": 1.6701648249891967, + "grad_norm": 0.21362553536891937, + "learning_rate": 2.6124585541486778e-06, + "loss": 0.0346, + "step": 54110 + }, + { + "epoch": 1.6704734860176553, + "grad_norm": 2.634674072265625, + "learning_rate": 2.6036744860974127e-06, + "loss": 0.0345, + "step": 54120 + }, + { + "epoch": 1.670782147046114, + "grad_norm": 0.29931363463401794, + "learning_rate": 2.594904815798399e-06, + "loss": 0.0361, + "step": 54130 + }, + { + "epoch": 1.6710908080745726, + "grad_norm": 0.34520840644836426, + "learning_rate": 2.586149545915656e-06, + "loss": 0.0324, + "step": 54140 + }, + { + "epoch": 1.6713994691030312, + "grad_norm": 0.31691330671310425, + "learning_rate": 2.577408679108778e-06, + "loss": 0.0341, + "step": 54150 + }, + { + "epoch": 1.6717081301314896, + "grad_norm": 0.26527655124664307, + "learning_rate": 2.5686822180330306e-06, + "loss": 0.0333, + "step": 54160 + }, + { + "epoch": 1.6720167911599482, + "grad_norm": 0.4899176359176636, + "learning_rate": 2.5599701653392703e-06, + "loss": 0.0326, + "step": 54170 + }, + { + "epoch": 1.6723254521884066, + "grad_norm": 0.26352161169052124, + "learning_rate": 2.551272523673992e-06, + "loss": 0.0324, + "step": 54180 + }, + { + "epoch": 1.6726341132168652, + "grad_norm": 0.37735211849212646, + "learning_rate": 2.542589295679315e-06, + "loss": 0.0327, + "step": 54190 + }, + { + "epoch": 1.6729427742453238, + "grad_norm": 0.3443707525730133, + "learning_rate": 2.5339204839929575e-06, + "loss": 0.0355, + "step": 54200 + }, + { + "epoch": 1.6732514352737824, + "grad_norm": 0.2805749475955963, + "learning_rate": 2.525266091248296e-06, + "loss": 0.034, + "step": 54210 + }, + { + "epoch": 1.673560096302241, + "grad_norm": 0.4206130802631378, + "learning_rate": 2.5166261200743e-06, + "loss": 0.0343, + "step": 54220 + }, + { + "epoch": 1.6738687573306994, + "grad_norm": 0.3308366537094116, + "learning_rate": 2.5080005730955646e-06, + "loss": 0.0324, + "step": 54230 + }, + { + "epoch": 1.674177418359158, + "grad_norm": 0.32907339930534363, + "learning_rate": 2.499389452932299e-06, + "loss": 0.038, + "step": 54240 + }, + { + "epoch": 1.6744860793876164, + "grad_norm": 0.3318629860877991, + "learning_rate": 2.4907927622003336e-06, + "loss": 0.0329, + "step": 54250 + }, + { + "epoch": 1.674794740416075, + "grad_norm": 0.3350474536418915, + "learning_rate": 2.4822105035111177e-06, + "loss": 0.0333, + "step": 54260 + }, + { + "epoch": 1.6751034014445336, + "grad_norm": 0.40109890699386597, + "learning_rate": 2.4736426794717273e-06, + "loss": 0.0336, + "step": 54270 + }, + { + "epoch": 1.6754120624729922, + "grad_norm": 0.44486841559410095, + "learning_rate": 2.4650892926848135e-06, + "loss": 0.0319, + "step": 54280 + }, + { + "epoch": 1.6757207235014508, + "grad_norm": 0.26352161169052124, + "learning_rate": 2.456550345748704e-06, + "loss": 0.0325, + "step": 54290 + }, + { + "epoch": 1.6760293845299092, + "grad_norm": 0.3376238942146301, + "learning_rate": 2.4480258412572733e-06, + "loss": 0.0337, + "step": 54300 + }, + { + "epoch": 1.6763380455583679, + "grad_norm": 0.5746715068817139, + "learning_rate": 2.4395157818000612e-06, + "loss": 0.0337, + "step": 54310 + }, + { + "epoch": 1.6766467065868262, + "grad_norm": 0.3416382670402527, + "learning_rate": 2.431020169962189e-06, + "loss": 0.0376, + "step": 54320 + }, + { + "epoch": 1.6769553676152849, + "grad_norm": 0.2752458155155182, + "learning_rate": 2.422539008324409e-06, + "loss": 0.0349, + "step": 54330 + }, + { + "epoch": 1.6772640286437435, + "grad_norm": 0.4721851944923401, + "learning_rate": 2.414072299463066e-06, + "loss": 0.0334, + "step": 54340 + }, + { + "epoch": 1.677572689672202, + "grad_norm": 0.4220113456249237, + "learning_rate": 2.4056200459501186e-06, + "loss": 0.0399, + "step": 54350 + }, + { + "epoch": 1.6778813507006607, + "grad_norm": 0.29099127650260925, + "learning_rate": 2.397182250353147e-06, + "loss": 0.0321, + "step": 54360 + }, + { + "epoch": 1.678190011729119, + "grad_norm": 0.3162844777107239, + "learning_rate": 2.388758915235334e-06, + "loss": 0.031, + "step": 54370 + }, + { + "epoch": 1.6784986727575775, + "grad_norm": 0.24295926094055176, + "learning_rate": 2.380350043155455e-06, + "loss": 0.0324, + "step": 54380 + }, + { + "epoch": 1.678807333786036, + "grad_norm": 0.2478037178516388, + "learning_rate": 2.371955636667911e-06, + "loss": 0.0321, + "step": 54390 + }, + { + "epoch": 1.6791159948144947, + "grad_norm": 0.6147136092185974, + "learning_rate": 2.3635756983227008e-06, + "loss": 0.0376, + "step": 54400 + }, + { + "epoch": 1.6794246558429533, + "grad_norm": 0.35256898403167725, + "learning_rate": 2.3552102306654278e-06, + "loss": 0.0338, + "step": 54410 + }, + { + "epoch": 1.679733316871412, + "grad_norm": 0.3933117091655731, + "learning_rate": 2.346859236237292e-06, + "loss": 0.0373, + "step": 54420 + }, + { + "epoch": 1.6800419778998703, + "grad_norm": 0.28110411763191223, + "learning_rate": 2.3385227175751145e-06, + "loss": 0.0298, + "step": 54430 + }, + { + "epoch": 1.680350638928329, + "grad_norm": 0.2772369384765625, + "learning_rate": 2.330200677211314e-06, + "loss": 0.0336, + "step": 54440 + }, + { + "epoch": 1.6806592999567873, + "grad_norm": 0.23731417953968048, + "learning_rate": 2.3218931176738847e-06, + "loss": 0.0361, + "step": 54450 + }, + { + "epoch": 1.680967960985246, + "grad_norm": 0.37748369574546814, + "learning_rate": 2.313600041486469e-06, + "loss": 0.0345, + "step": 54460 + }, + { + "epoch": 1.6812766220137045, + "grad_norm": 0.3770034611225128, + "learning_rate": 2.3053214511682743e-06, + "loss": 0.0353, + "step": 54470 + }, + { + "epoch": 1.6815852830421631, + "grad_norm": 0.32557806372642517, + "learning_rate": 2.2970573492341163e-06, + "loss": 0.0341, + "step": 54480 + }, + { + "epoch": 1.6818939440706218, + "grad_norm": 0.3227680027484894, + "learning_rate": 2.288807738194415e-06, + "loss": 0.034, + "step": 54490 + }, + { + "epoch": 1.6822026050990802, + "grad_norm": 0.32403844594955444, + "learning_rate": 2.2805726205551768e-06, + "loss": 0.0312, + "step": 54500 + }, + { + "epoch": 1.6825112661275388, + "grad_norm": 0.3821282684803009, + "learning_rate": 2.272351998818023e-06, + "loss": 0.0344, + "step": 54510 + }, + { + "epoch": 1.6828199271559972, + "grad_norm": 0.28996455669403076, + "learning_rate": 2.2641458754801505e-06, + "loss": 0.0299, + "step": 54520 + }, + { + "epoch": 1.6831285881844558, + "grad_norm": 0.26980406045913696, + "learning_rate": 2.2559542530343756e-06, + "loss": 0.0321, + "step": 54530 + }, + { + "epoch": 1.6834372492129144, + "grad_norm": 0.2699815034866333, + "learning_rate": 2.247777133969087e-06, + "loss": 0.0326, + "step": 54540 + }, + { + "epoch": 1.683745910241373, + "grad_norm": 0.20820701122283936, + "learning_rate": 2.2396145207682795e-06, + "loss": 0.0333, + "step": 54550 + }, + { + "epoch": 1.6840545712698316, + "grad_norm": 0.34425729513168335, + "learning_rate": 2.231466415911543e-06, + "loss": 0.0368, + "step": 54560 + }, + { + "epoch": 1.68436323229829, + "grad_norm": 0.4234224259853363, + "learning_rate": 2.2233328218740524e-06, + "loss": 0.034, + "step": 54570 + }, + { + "epoch": 1.6846718933267486, + "grad_norm": 0.6175723075866699, + "learning_rate": 2.215213741126576e-06, + "loss": 0.0341, + "step": 54580 + }, + { + "epoch": 1.684980554355207, + "grad_norm": 0.3270523250102997, + "learning_rate": 2.2071091761354912e-06, + "loss": 0.0328, + "step": 54590 + }, + { + "epoch": 1.6852892153836656, + "grad_norm": 0.41098907589912415, + "learning_rate": 2.1990191293627337e-06, + "loss": 0.036, + "step": 54600 + }, + { + "epoch": 1.6855978764121242, + "grad_norm": 0.35459741950035095, + "learning_rate": 2.1909436032658548e-06, + "loss": 0.0334, + "step": 54610 + }, + { + "epoch": 1.6859065374405828, + "grad_norm": 0.2903318703174591, + "learning_rate": 2.1828826002979806e-06, + "loss": 0.0349, + "step": 54620 + }, + { + "epoch": 1.6862151984690414, + "grad_norm": 0.2967153489589691, + "learning_rate": 2.1748361229078362e-06, + "loss": 0.0328, + "step": 54630 + }, + { + "epoch": 1.6865238594974998, + "grad_norm": 0.23317131400108337, + "learning_rate": 2.1668041735397327e-06, + "loss": 0.0336, + "step": 54640 + }, + { + "epoch": 1.6868325205259584, + "grad_norm": 0.3851669728755951, + "learning_rate": 2.1587867546335514e-06, + "loss": 0.0354, + "step": 54650 + }, + { + "epoch": 1.6871411815544168, + "grad_norm": 0.27975940704345703, + "learning_rate": 2.1507838686247894e-06, + "loss": 0.0348, + "step": 54660 + }, + { + "epoch": 1.6874498425828754, + "grad_norm": 0.44569218158721924, + "learning_rate": 2.1427955179444957e-06, + "loss": 0.0316, + "step": 54670 + }, + { + "epoch": 1.687758503611334, + "grad_norm": 0.27690133452415466, + "learning_rate": 2.13482170501933e-06, + "loss": 0.036, + "step": 54680 + }, + { + "epoch": 1.6880671646397927, + "grad_norm": 0.31630200147628784, + "learning_rate": 2.1268624322715202e-06, + "loss": 0.0328, + "step": 54690 + }, + { + "epoch": 1.6883758256682513, + "grad_norm": 0.305963933467865, + "learning_rate": 2.118917702118889e-06, + "loss": 0.0307, + "step": 54700 + }, + { + "epoch": 1.6886844866967097, + "grad_norm": 0.27952924370765686, + "learning_rate": 2.1109875169748327e-06, + "loss": 0.0315, + "step": 54710 + }, + { + "epoch": 1.688993147725168, + "grad_norm": 0.5964692831039429, + "learning_rate": 2.103071879248336e-06, + "loss": 0.0298, + "step": 54720 + }, + { + "epoch": 1.6893018087536267, + "grad_norm": 0.422721803188324, + "learning_rate": 2.0951707913439478e-06, + "loss": 0.0367, + "step": 54730 + }, + { + "epoch": 1.6896104697820853, + "grad_norm": 0.26506227254867554, + "learning_rate": 2.0872842556618255e-06, + "loss": 0.0322, + "step": 54740 + }, + { + "epoch": 1.689919130810544, + "grad_norm": 0.34477925300598145, + "learning_rate": 2.079412274597686e-06, + "loss": 0.0317, + "step": 54750 + }, + { + "epoch": 1.6902277918390025, + "grad_norm": 0.25657418370246887, + "learning_rate": 2.0715548505428284e-06, + "loss": 0.0389, + "step": 54760 + }, + { + "epoch": 1.690536452867461, + "grad_norm": 0.37573209404945374, + "learning_rate": 2.0637119858841258e-06, + "loss": 0.0315, + "step": 54770 + }, + { + "epoch": 1.6908451138959195, + "grad_norm": 0.2995377779006958, + "learning_rate": 2.055883683004034e-06, + "loss": 0.0386, + "step": 54780 + }, + { + "epoch": 1.691153774924378, + "grad_norm": 0.5431077480316162, + "learning_rate": 2.0480699442806006e-06, + "loss": 0.0303, + "step": 54790 + }, + { + "epoch": 1.6914624359528365, + "grad_norm": 0.4411489963531494, + "learning_rate": 2.0402707720874105e-06, + "loss": 0.0315, + "step": 54800 + }, + { + "epoch": 1.6917710969812951, + "grad_norm": 0.3984162509441376, + "learning_rate": 2.032486168793668e-06, + "loss": 0.0363, + "step": 54810 + }, + { + "epoch": 1.6920797580097537, + "grad_norm": 0.2737583816051483, + "learning_rate": 2.024716136764104e-06, + "loss": 0.0332, + "step": 54820 + }, + { + "epoch": 1.6923884190382124, + "grad_norm": 0.37670204043388367, + "learning_rate": 2.0169606783590757e-06, + "loss": 0.0364, + "step": 54830 + }, + { + "epoch": 1.6926970800666707, + "grad_norm": 0.31203946471214294, + "learning_rate": 2.0092197959344638e-06, + "loss": 0.0325, + "step": 54840 + }, + { + "epoch": 1.6930057410951294, + "grad_norm": 0.421124130487442, + "learning_rate": 2.0014934918417606e-06, + "loss": 0.0338, + "step": 54850 + }, + { + "epoch": 1.6933144021235877, + "grad_norm": 0.4053633213043213, + "learning_rate": 1.99378176842801e-06, + "loss": 0.0336, + "step": 54860 + }, + { + "epoch": 1.6936230631520464, + "grad_norm": 0.39247140288352966, + "learning_rate": 1.9860846280358224e-06, + "loss": 0.0327, + "step": 54870 + }, + { + "epoch": 1.693931724180505, + "grad_norm": 0.37567317485809326, + "learning_rate": 1.978402073003394e-06, + "loss": 0.0328, + "step": 54880 + }, + { + "epoch": 1.6942403852089636, + "grad_norm": 0.3732362687587738, + "learning_rate": 1.9707341056644736e-06, + "loss": 0.0383, + "step": 54890 + }, + { + "epoch": 1.6945490462374222, + "grad_norm": 0.5442652702331543, + "learning_rate": 1.963080728348399e-06, + "loss": 0.0353, + "step": 54900 + }, + { + "epoch": 1.6948577072658806, + "grad_norm": 0.4422760307788849, + "learning_rate": 1.955441943380054e-06, + "loss": 0.029, + "step": 54910 + }, + { + "epoch": 1.6951663682943392, + "grad_norm": 0.30838701128959656, + "learning_rate": 1.947817753079906e-06, + "loss": 0.0353, + "step": 54920 + }, + { + "epoch": 1.6954750293227976, + "grad_norm": 0.2872578799724579, + "learning_rate": 1.9402081597639785e-06, + "loss": 0.0322, + "step": 54930 + }, + { + "epoch": 1.6957836903512562, + "grad_norm": 0.35476914048194885, + "learning_rate": 1.9326131657438683e-06, + "loss": 0.0319, + "step": 54940 + }, + { + "epoch": 1.6960923513797148, + "grad_norm": 0.3963032364845276, + "learning_rate": 1.925032773326724e-06, + "loss": 0.0391, + "step": 54950 + }, + { + "epoch": 1.6964010124081734, + "grad_norm": 0.3845367431640625, + "learning_rate": 1.9174669848152916e-06, + "loss": 0.0328, + "step": 54960 + }, + { + "epoch": 1.696709673436632, + "grad_norm": 0.3303926885128021, + "learning_rate": 1.9099158025078336e-06, + "loss": 0.0314, + "step": 54970 + }, + { + "epoch": 1.6970183344650904, + "grad_norm": 0.28841617703437805, + "learning_rate": 1.90237922869822e-06, + "loss": 0.0334, + "step": 54980 + }, + { + "epoch": 1.697326995493549, + "grad_norm": 0.3260829448699951, + "learning_rate": 1.8948572656758367e-06, + "loss": 0.0291, + "step": 54990 + }, + { + "epoch": 1.6976356565220074, + "grad_norm": 0.30249717831611633, + "learning_rate": 1.8873499157256834e-06, + "loss": 0.0358, + "step": 55000 + }, + { + "epoch": 1.697944317550466, + "grad_norm": 0.32429227232933044, + "learning_rate": 1.879857181128286e-06, + "loss": 0.0304, + "step": 55010 + }, + { + "epoch": 1.6982529785789247, + "grad_norm": 0.3590320646762848, + "learning_rate": 1.8723790641597349e-06, + "loss": 0.0368, + "step": 55020 + }, + { + "epoch": 1.6985616396073833, + "grad_norm": 0.28133660554885864, + "learning_rate": 1.8649155670916906e-06, + "loss": 0.0329, + "step": 55030 + }, + { + "epoch": 1.6988703006358419, + "grad_norm": 0.2870931029319763, + "learning_rate": 1.8574666921913565e-06, + "loss": 0.0371, + "step": 55040 + }, + { + "epoch": 1.6991789616643003, + "grad_norm": 0.3607645034790039, + "learning_rate": 1.8500324417215166e-06, + "loss": 0.0331, + "step": 55050 + }, + { + "epoch": 1.6994876226927587, + "grad_norm": 0.3541082441806793, + "learning_rate": 1.8426128179404977e-06, + "loss": 0.031, + "step": 55060 + }, + { + "epoch": 1.6997962837212173, + "grad_norm": 0.33945825695991516, + "learning_rate": 1.8352078231021807e-06, + "loss": 0.0348, + "step": 55070 + }, + { + "epoch": 1.7001049447496759, + "grad_norm": 0.5203740000724792, + "learning_rate": 1.8278174594560049e-06, + "loss": 0.0333, + "step": 55080 + }, + { + "epoch": 1.7004136057781345, + "grad_norm": 0.2825639843940735, + "learning_rate": 1.82044172924698e-06, + "loss": 0.0336, + "step": 55090 + }, + { + "epoch": 1.700722266806593, + "grad_norm": 0.2962706983089447, + "learning_rate": 1.813080634715636e-06, + "loss": 0.0322, + "step": 55100 + }, + { + "epoch": 1.7010309278350515, + "grad_norm": 0.36519911885261536, + "learning_rate": 1.8057341780981119e-06, + "loss": 0.0344, + "step": 55110 + }, + { + "epoch": 1.70133958886351, + "grad_norm": 0.3470930755138397, + "learning_rate": 1.7984023616260338e-06, + "loss": 0.0376, + "step": 55120 + }, + { + "epoch": 1.7016482498919685, + "grad_norm": 0.3327794671058655, + "learning_rate": 1.7910851875266421e-06, + "loss": 0.0294, + "step": 55130 + }, + { + "epoch": 1.7019569109204271, + "grad_norm": 0.2758137881755829, + "learning_rate": 1.7837826580226757e-06, + "loss": 0.0334, + "step": 55140 + }, + { + "epoch": 1.7022655719488857, + "grad_norm": 0.5214805006980896, + "learning_rate": 1.7764947753324656e-06, + "loss": 0.0334, + "step": 55150 + }, + { + "epoch": 1.7025742329773443, + "grad_norm": 0.32058489322662354, + "learning_rate": 1.7692215416698799e-06, + "loss": 0.0295, + "step": 55160 + }, + { + "epoch": 1.702882894005803, + "grad_norm": 0.3399323523044586, + "learning_rate": 1.7619629592443233e-06, + "loss": 0.0357, + "step": 55170 + }, + { + "epoch": 1.7031915550342613, + "grad_norm": 0.36662423610687256, + "learning_rate": 1.7547190302607709e-06, + "loss": 0.0334, + "step": 55180 + }, + { + "epoch": 1.70350021606272, + "grad_norm": 0.3904033899307251, + "learning_rate": 1.747489756919729e-06, + "loss": 0.0352, + "step": 55190 + }, + { + "epoch": 1.7038088770911783, + "grad_norm": 0.5025049448013306, + "learning_rate": 1.7402751414172802e-06, + "loss": 0.0372, + "step": 55200 + }, + { + "epoch": 1.704117538119637, + "grad_norm": 0.36837607622146606, + "learning_rate": 1.7330751859450044e-06, + "loss": 0.0299, + "step": 55210 + }, + { + "epoch": 1.7044261991480956, + "grad_norm": 0.2669804096221924, + "learning_rate": 1.7258898926900801e-06, + "loss": 0.0318, + "step": 55220 + }, + { + "epoch": 1.7047348601765542, + "grad_norm": 0.3037330210208893, + "learning_rate": 1.7187192638352002e-06, + "loss": 0.0314, + "step": 55230 + }, + { + "epoch": 1.7050435212050128, + "grad_norm": 0.6541587710380554, + "learning_rate": 1.7115633015586163e-06, + "loss": 0.0322, + "step": 55240 + }, + { + "epoch": 1.7053521822334712, + "grad_norm": 0.35481253266334534, + "learning_rate": 1.7044220080341178e-06, + "loss": 0.0354, + "step": 55250 + }, + { + "epoch": 1.7056608432619298, + "grad_norm": 0.4531717896461487, + "learning_rate": 1.6972953854310414e-06, + "loss": 0.0325, + "step": 55260 + }, + { + "epoch": 1.7059695042903882, + "grad_norm": 0.328477680683136, + "learning_rate": 1.690183435914261e-06, + "loss": 0.031, + "step": 55270 + }, + { + "epoch": 1.7062781653188468, + "grad_norm": 0.24226966500282288, + "learning_rate": 1.6830861616442206e-06, + "loss": 0.0347, + "step": 55280 + }, + { + "epoch": 1.7065868263473054, + "grad_norm": 0.344673216342926, + "learning_rate": 1.6760035647768568e-06, + "loss": 0.0334, + "step": 55290 + }, + { + "epoch": 1.706895487375764, + "grad_norm": 0.33921757340431213, + "learning_rate": 1.6689356474636985e-06, + "loss": 0.037, + "step": 55300 + }, + { + "epoch": 1.7072041484042226, + "grad_norm": 0.665942370891571, + "learning_rate": 1.6618824118517784e-06, + "loss": 0.0357, + "step": 55310 + }, + { + "epoch": 1.707512809432681, + "grad_norm": 0.43694013357162476, + "learning_rate": 1.654843860083688e-06, + "loss": 0.0311, + "step": 55320 + }, + { + "epoch": 1.7078214704611394, + "grad_norm": 0.3650432825088501, + "learning_rate": 1.647819994297556e-06, + "loss": 0.0332, + "step": 55330 + }, + { + "epoch": 1.708130131489598, + "grad_norm": 0.31208890676498413, + "learning_rate": 1.6408108166270363e-06, + "loss": 0.0325, + "step": 55340 + }, + { + "epoch": 1.7084387925180566, + "grad_norm": 0.27373650670051575, + "learning_rate": 1.6338163292013486e-06, + "loss": 0.0328, + "step": 55350 + }, + { + "epoch": 1.7087474535465152, + "grad_norm": 0.35389435291290283, + "learning_rate": 1.6268365341452208e-06, + "loss": 0.0333, + "step": 55360 + }, + { + "epoch": 1.7090561145749739, + "grad_norm": 0.3154619634151459, + "learning_rate": 1.6198714335789345e-06, + "loss": 0.0345, + "step": 55370 + }, + { + "epoch": 1.7093647756034325, + "grad_norm": 0.31154778599739075, + "learning_rate": 1.612921029618303e-06, + "loss": 0.0355, + "step": 55380 + }, + { + "epoch": 1.7096734366318909, + "grad_norm": 0.2639178931713104, + "learning_rate": 1.6059853243746815e-06, + "loss": 0.0353, + "step": 55390 + }, + { + "epoch": 1.7099820976603493, + "grad_norm": 0.32540059089660645, + "learning_rate": 1.5990643199549404e-06, + "loss": 0.0351, + "step": 55400 + }, + { + "epoch": 1.7102907586888079, + "grad_norm": 0.2958686351776123, + "learning_rate": 1.5921580184615147e-06, + "loss": 0.0329, + "step": 55410 + }, + { + "epoch": 1.7105994197172665, + "grad_norm": 0.36926552653312683, + "learning_rate": 1.5852664219923374e-06, + "loss": 0.0321, + "step": 55420 + }, + { + "epoch": 1.710908080745725, + "grad_norm": 0.35560163855552673, + "learning_rate": 1.5783895326409172e-06, + "loss": 0.0323, + "step": 55430 + }, + { + "epoch": 1.7112167417741837, + "grad_norm": 0.33736690878868103, + "learning_rate": 1.5715273524962438e-06, + "loss": 0.0302, + "step": 55440 + }, + { + "epoch": 1.711525402802642, + "grad_norm": 0.32812198996543884, + "learning_rate": 1.564679883642889e-06, + "loss": 0.0307, + "step": 55450 + }, + { + "epoch": 1.7118340638311007, + "grad_norm": 0.3558453619480133, + "learning_rate": 1.5578471281609276e-06, + "loss": 0.0336, + "step": 55460 + }, + { + "epoch": 1.712142724859559, + "grad_norm": 0.3259905278682709, + "learning_rate": 1.551029088125966e-06, + "loss": 0.0332, + "step": 55470 + }, + { + "epoch": 1.7124513858880177, + "grad_norm": 0.31664958596229553, + "learning_rate": 1.544225765609142e-06, + "loss": 0.0313, + "step": 55480 + }, + { + "epoch": 1.7127600469164763, + "grad_norm": 0.4031144380569458, + "learning_rate": 1.53743716267713e-06, + "loss": 0.0339, + "step": 55490 + }, + { + "epoch": 1.713068707944935, + "grad_norm": 0.5014376640319824, + "learning_rate": 1.5306632813921361e-06, + "loss": 0.034, + "step": 55500 + }, + { + "epoch": 1.7133773689733935, + "grad_norm": 0.32045990228652954, + "learning_rate": 1.52390412381187e-06, + "loss": 0.0344, + "step": 55510 + }, + { + "epoch": 1.713686030001852, + "grad_norm": 0.31576696038246155, + "learning_rate": 1.5171596919895948e-06, + "loss": 0.0344, + "step": 55520 + }, + { + "epoch": 1.7139946910303105, + "grad_norm": 0.2670336365699768, + "learning_rate": 1.5104299879740936e-06, + "loss": 0.0327, + "step": 55530 + }, + { + "epoch": 1.714303352058769, + "grad_norm": 0.44415929913520813, + "learning_rate": 1.5037150138096701e-06, + "loss": 0.0343, + "step": 55540 + }, + { + "epoch": 1.7146120130872275, + "grad_norm": 0.31453025341033936, + "learning_rate": 1.4970147715361538e-06, + "loss": 0.0304, + "step": 55550 + }, + { + "epoch": 1.7149206741156862, + "grad_norm": 0.3424915075302124, + "learning_rate": 1.4903292631889054e-06, + "loss": 0.0329, + "step": 55560 + }, + { + "epoch": 1.7152293351441448, + "grad_norm": 0.3072352707386017, + "learning_rate": 1.483658490798795e-06, + "loss": 0.0365, + "step": 55570 + }, + { + "epoch": 1.7155379961726034, + "grad_norm": 0.3760654926300049, + "learning_rate": 1.4770024563922457e-06, + "loss": 0.0326, + "step": 55580 + }, + { + "epoch": 1.7158466572010618, + "grad_norm": 0.28060948848724365, + "learning_rate": 1.4703611619911628e-06, + "loss": 0.0344, + "step": 55590 + }, + { + "epoch": 1.7161553182295204, + "grad_norm": 0.35984399914741516, + "learning_rate": 1.4637346096130155e-06, + "loss": 0.0336, + "step": 55600 + }, + { + "epoch": 1.7164639792579788, + "grad_norm": 0.3804537355899811, + "learning_rate": 1.4571228012707662e-06, + "loss": 0.0371, + "step": 55610 + }, + { + "epoch": 1.7167726402864374, + "grad_norm": 0.32169026136398315, + "learning_rate": 1.4505257389729132e-06, + "loss": 0.0379, + "step": 55620 + }, + { + "epoch": 1.717081301314896, + "grad_norm": 0.31144043803215027, + "learning_rate": 1.4439434247234596e-06, + "loss": 0.0339, + "step": 55630 + }, + { + "epoch": 1.7173899623433546, + "grad_norm": 0.33462491631507874, + "learning_rate": 1.4373758605219445e-06, + "loss": 0.0314, + "step": 55640 + }, + { + "epoch": 1.7176986233718132, + "grad_norm": 0.4496821463108063, + "learning_rate": 1.4308230483634333e-06, + "loss": 0.0357, + "step": 55650 + }, + { + "epoch": 1.7180072844002716, + "grad_norm": 0.24208283424377441, + "learning_rate": 1.4242849902384724e-06, + "loss": 0.0306, + "step": 55660 + }, + { + "epoch": 1.71831594542873, + "grad_norm": 0.44212228059768677, + "learning_rate": 1.4177616881331734e-06, + "loss": 0.0323, + "step": 55670 + }, + { + "epoch": 1.7186246064571886, + "grad_norm": 0.3312666118144989, + "learning_rate": 1.4112531440291233e-06, + "loss": 0.0349, + "step": 55680 + }, + { + "epoch": 1.7189332674856472, + "grad_norm": 0.25469645857810974, + "learning_rate": 1.4047593599034624e-06, + "loss": 0.0346, + "step": 55690 + }, + { + "epoch": 1.7192419285141058, + "grad_norm": 0.3942919969558716, + "learning_rate": 1.3982803377288246e-06, + "loss": 0.0321, + "step": 55700 + }, + { + "epoch": 1.7195505895425645, + "grad_norm": 0.28442057967185974, + "learning_rate": 1.3918160794733681e-06, + "loss": 0.0333, + "step": 55710 + }, + { + "epoch": 1.719859250571023, + "grad_norm": 0.3691231608390808, + "learning_rate": 1.3853665871007615e-06, + "loss": 0.0326, + "step": 55720 + }, + { + "epoch": 1.7201679115994815, + "grad_norm": 0.31005680561065674, + "learning_rate": 1.378931862570193e-06, + "loss": 0.0345, + "step": 55730 + }, + { + "epoch": 1.7204765726279398, + "grad_norm": 0.372916042804718, + "learning_rate": 1.372511907836349e-06, + "loss": 0.0334, + "step": 55740 + }, + { + "epoch": 1.7207852336563985, + "grad_norm": 0.38029250502586365, + "learning_rate": 1.3661067248494586e-06, + "loss": 0.0363, + "step": 55750 + }, + { + "epoch": 1.721093894684857, + "grad_norm": 0.4106735587120056, + "learning_rate": 1.3597163155552429e-06, + "loss": 0.0309, + "step": 55760 + }, + { + "epoch": 1.7214025557133157, + "grad_norm": 0.4605600833892822, + "learning_rate": 1.3533406818949434e-06, + "loss": 0.0355, + "step": 55770 + }, + { + "epoch": 1.7217112167417743, + "grad_norm": 0.5868194103240967, + "learning_rate": 1.3469798258053002e-06, + "loss": 0.032, + "step": 55780 + }, + { + "epoch": 1.7220198777702327, + "grad_norm": 0.35499122738838196, + "learning_rate": 1.3406337492185672e-06, + "loss": 0.0342, + "step": 55790 + }, + { + "epoch": 1.7223285387986913, + "grad_norm": 0.5842518210411072, + "learning_rate": 1.3343024540625414e-06, + "loss": 0.0335, + "step": 55800 + }, + { + "epoch": 1.7226371998271497, + "grad_norm": 0.25907039642333984, + "learning_rate": 1.3279859422604735e-06, + "loss": 0.0331, + "step": 55810 + }, + { + "epoch": 1.7229458608556083, + "grad_norm": 0.38750723004341125, + "learning_rate": 1.3216842157311781e-06, + "loss": 0.0333, + "step": 55820 + }, + { + "epoch": 1.723254521884067, + "grad_norm": 0.29316917061805725, + "learning_rate": 1.3153972763889355e-06, + "loss": 0.0346, + "step": 55830 + }, + { + "epoch": 1.7235631829125255, + "grad_norm": 0.4355204999446869, + "learning_rate": 1.3091251261435566e-06, + "loss": 0.0318, + "step": 55840 + }, + { + "epoch": 1.7238718439409841, + "grad_norm": 0.42665815353393555, + "learning_rate": 1.3028677669003564e-06, + "loss": 0.0334, + "step": 55850 + }, + { + "epoch": 1.7241805049694425, + "grad_norm": 0.3876821994781494, + "learning_rate": 1.2966252005601587e-06, + "loss": 0.0345, + "step": 55860 + }, + { + "epoch": 1.7244891659979011, + "grad_norm": 0.2477392703294754, + "learning_rate": 1.2903974290192855e-06, + "loss": 0.0328, + "step": 55870 + }, + { + "epoch": 1.7247978270263595, + "grad_norm": 0.5111454129219055, + "learning_rate": 1.284184454169568e-06, + "loss": 0.0322, + "step": 55880 + }, + { + "epoch": 1.7251064880548181, + "grad_norm": 0.25682032108306885, + "learning_rate": 1.2779862778983464e-06, + "loss": 0.032, + "step": 55890 + }, + { + "epoch": 1.7254151490832768, + "grad_norm": 0.3843757212162018, + "learning_rate": 1.2718029020884647e-06, + "loss": 0.0324, + "step": 55900 + }, + { + "epoch": 1.7257238101117354, + "grad_norm": 0.3537176847457886, + "learning_rate": 1.2656343286182703e-06, + "loss": 0.0326, + "step": 55910 + }, + { + "epoch": 1.726032471140194, + "grad_norm": 0.28171056509017944, + "learning_rate": 1.2594805593616088e-06, + "loss": 0.0334, + "step": 55920 + }, + { + "epoch": 1.7263411321686524, + "grad_norm": 0.3596676290035248, + "learning_rate": 1.2533415961878404e-06, + "loss": 0.0361, + "step": 55930 + }, + { + "epoch": 1.726649793197111, + "grad_norm": 0.3330429196357727, + "learning_rate": 1.2472174409618009e-06, + "loss": 0.0359, + "step": 55940 + }, + { + "epoch": 1.7269584542255694, + "grad_norm": 0.3241024315357208, + "learning_rate": 1.2411080955438747e-06, + "loss": 0.0347, + "step": 55950 + }, + { + "epoch": 1.727267115254028, + "grad_norm": 0.3329354226589203, + "learning_rate": 1.235013561789894e-06, + "loss": 0.0354, + "step": 55960 + }, + { + "epoch": 1.7275757762824866, + "grad_norm": 0.2327669858932495, + "learning_rate": 1.2289338415512385e-06, + "loss": 0.0295, + "step": 55970 + }, + { + "epoch": 1.7278844373109452, + "grad_norm": 0.3404938280582428, + "learning_rate": 1.222868936674748e-06, + "loss": 0.0328, + "step": 55980 + }, + { + "epoch": 1.7281930983394038, + "grad_norm": 0.28333666920661926, + "learning_rate": 1.2168188490027876e-06, + "loss": 0.0355, + "step": 55990 + }, + { + "epoch": 1.7285017593678622, + "grad_norm": 0.28693220019340515, + "learning_rate": 1.2107835803732204e-06, + "loss": 0.0356, + "step": 56000 + }, + { + "epoch": 1.7288104203963206, + "grad_norm": 0.28291556239128113, + "learning_rate": 1.2047631326193964e-06, + "loss": 0.0343, + "step": 56010 + }, + { + "epoch": 1.7291190814247792, + "grad_norm": 0.24689984321594238, + "learning_rate": 1.1987575075701696e-06, + "loss": 0.034, + "step": 56020 + }, + { + "epoch": 1.7294277424532378, + "grad_norm": 0.3574024736881256, + "learning_rate": 1.1927667070498916e-06, + "loss": 0.0317, + "step": 56030 + }, + { + "epoch": 1.7297364034816964, + "grad_norm": 0.2695523202419281, + "learning_rate": 1.1867907328784067e-06, + "loss": 0.0394, + "step": 56040 + }, + { + "epoch": 1.730045064510155, + "grad_norm": 0.3045026659965515, + "learning_rate": 1.1808295868710518e-06, + "loss": 0.0355, + "step": 56050 + }, + { + "epoch": 1.7303537255386134, + "grad_norm": 0.23184603452682495, + "learning_rate": 1.174883270838678e-06, + "loss": 0.0348, + "step": 56060 + }, + { + "epoch": 1.730662386567072, + "grad_norm": 0.3226670026779175, + "learning_rate": 1.1689517865876187e-06, + "loss": 0.0372, + "step": 56070 + }, + { + "epoch": 1.7309710475955304, + "grad_norm": 0.33854153752326965, + "learning_rate": 1.1630351359196933e-06, + "loss": 0.0293, + "step": 56080 + }, + { + "epoch": 1.731279708623989, + "grad_norm": 0.2751368582248688, + "learning_rate": 1.1571333206322255e-06, + "loss": 0.0336, + "step": 56090 + }, + { + "epoch": 1.7315883696524477, + "grad_norm": 0.33865833282470703, + "learning_rate": 1.1512463425180365e-06, + "loss": 0.0334, + "step": 56100 + }, + { + "epoch": 1.7318970306809063, + "grad_norm": 0.2647077441215515, + "learning_rate": 1.145374203365429e-06, + "loss": 0.037, + "step": 56110 + }, + { + "epoch": 1.732205691709365, + "grad_norm": 0.30288979411125183, + "learning_rate": 1.1395169049582155e-06, + "loss": 0.0307, + "step": 56120 + }, + { + "epoch": 1.7325143527378233, + "grad_norm": 0.23799417912960052, + "learning_rate": 1.1336744490756722e-06, + "loss": 0.0364, + "step": 56130 + }, + { + "epoch": 1.732823013766282, + "grad_norm": 0.4338604509830475, + "learning_rate": 1.1278468374925967e-06, + "loss": 0.0356, + "step": 56140 + }, + { + "epoch": 1.7331316747947403, + "grad_norm": 0.6161218881607056, + "learning_rate": 1.12203407197925e-06, + "loss": 0.0357, + "step": 56150 + }, + { + "epoch": 1.733440335823199, + "grad_norm": 0.3439195156097412, + "learning_rate": 1.116236154301409e-06, + "loss": 0.0323, + "step": 56160 + }, + { + "epoch": 1.7337489968516575, + "grad_norm": 0.29469314217567444, + "learning_rate": 1.11045308622032e-06, + "loss": 0.0317, + "step": 56170 + }, + { + "epoch": 1.7340576578801161, + "grad_norm": 0.3868810534477234, + "learning_rate": 1.1046848694927337e-06, + "loss": 0.0377, + "step": 56180 + }, + { + "epoch": 1.7343663189085747, + "grad_norm": 0.3540664613246918, + "learning_rate": 1.098931505870876e-06, + "loss": 0.0316, + "step": 56190 + }, + { + "epoch": 1.7346749799370331, + "grad_norm": 0.6391710042953491, + "learning_rate": 1.0931929971024657e-06, + "loss": 0.0338, + "step": 56200 + }, + { + "epoch": 1.7349836409654917, + "grad_norm": 0.3294770121574402, + "learning_rate": 1.0874693449307193e-06, + "loss": 0.0342, + "step": 56210 + }, + { + "epoch": 1.7352923019939501, + "grad_norm": 0.47640129923820496, + "learning_rate": 1.081760551094324e-06, + "loss": 0.04, + "step": 56220 + }, + { + "epoch": 1.7356009630224087, + "grad_norm": 0.2659584879875183, + "learning_rate": 1.0760666173274592e-06, + "loss": 0.0339, + "step": 56230 + }, + { + "epoch": 1.7359096240508674, + "grad_norm": 0.5473900437355042, + "learning_rate": 1.0703875453597967e-06, + "loss": 0.0306, + "step": 56240 + }, + { + "epoch": 1.736218285079326, + "grad_norm": 0.3455555737018585, + "learning_rate": 1.0647233369164845e-06, + "loss": 0.034, + "step": 56250 + }, + { + "epoch": 1.7365269461077846, + "grad_norm": 0.4393349587917328, + "learning_rate": 1.0590739937181625e-06, + "loss": 0.0369, + "step": 56260 + }, + { + "epoch": 1.736835607136243, + "grad_norm": 0.31149613857269287, + "learning_rate": 1.053439517480953e-06, + "loss": 0.0356, + "step": 56270 + }, + { + "epoch": 1.7371442681647016, + "grad_norm": 0.3058181703090668, + "learning_rate": 1.047819909916453e-06, + "loss": 0.0335, + "step": 56280 + }, + { + "epoch": 1.73745292919316, + "grad_norm": 0.34310027956962585, + "learning_rate": 1.0422151727317697e-06, + "loss": 0.0339, + "step": 56290 + }, + { + "epoch": 1.7377615902216186, + "grad_norm": 0.30054646730422974, + "learning_rate": 1.0366253076294462e-06, + "loss": 0.032, + "step": 56300 + }, + { + "epoch": 1.7380702512500772, + "grad_norm": 0.3227993845939636, + "learning_rate": 1.031050316307558e-06, + "loss": 0.0373, + "step": 56310 + }, + { + "epoch": 1.7383789122785358, + "grad_norm": 0.44523632526397705, + "learning_rate": 1.0254902004596333e-06, + "loss": 0.0394, + "step": 56320 + }, + { + "epoch": 1.7386875733069944, + "grad_norm": 0.3508768081665039, + "learning_rate": 1.0199449617746882e-06, + "loss": 0.0328, + "step": 56330 + }, + { + "epoch": 1.7389962343354528, + "grad_norm": 0.29521238803863525, + "learning_rate": 1.0144146019372247e-06, + "loss": 0.0326, + "step": 56340 + }, + { + "epoch": 1.7393048953639112, + "grad_norm": 0.36048224568367004, + "learning_rate": 1.0088991226272048e-06, + "loss": 0.0359, + "step": 56350 + }, + { + "epoch": 1.7396135563923698, + "grad_norm": 0.3020803928375244, + "learning_rate": 1.003398525520105e-06, + "loss": 0.0398, + "step": 56360 + }, + { + "epoch": 1.7399222174208284, + "grad_norm": 0.2753317654132843, + "learning_rate": 9.979128122868552e-07, + "loss": 0.0345, + "step": 56370 + }, + { + "epoch": 1.740230878449287, + "grad_norm": 0.3468377888202667, + "learning_rate": 9.924419845938614e-07, + "loss": 0.0313, + "step": 56380 + }, + { + "epoch": 1.7405395394777456, + "grad_norm": 0.3606873154640198, + "learning_rate": 9.869860441030276e-07, + "loss": 0.0364, + "step": 56390 + }, + { + "epoch": 1.740848200506204, + "grad_norm": 0.5311044454574585, + "learning_rate": 9.815449924717169e-07, + "loss": 0.036, + "step": 56400 + }, + { + "epoch": 1.7411568615346626, + "grad_norm": 0.32693251967430115, + "learning_rate": 9.761188313527791e-07, + "loss": 0.0313, + "step": 56410 + }, + { + "epoch": 1.741465522563121, + "grad_norm": 0.29289525747299194, + "learning_rate": 9.70707562394546e-07, + "loss": 0.0341, + "step": 56420 + }, + { + "epoch": 1.7417741835915796, + "grad_norm": 0.4616764783859253, + "learning_rate": 9.653111872408027e-07, + "loss": 0.0372, + "step": 56430 + }, + { + "epoch": 1.7420828446200383, + "grad_norm": 0.3655655086040497, + "learning_rate": 9.599297075308434e-07, + "loss": 0.0334, + "step": 56440 + }, + { + "epoch": 1.7423915056484969, + "grad_norm": 0.24541506171226501, + "learning_rate": 9.545631248994048e-07, + "loss": 0.0328, + "step": 56450 + }, + { + "epoch": 1.7427001666769555, + "grad_norm": 0.21986275911331177, + "learning_rate": 9.492114409767217e-07, + "loss": 0.0318, + "step": 56460 + }, + { + "epoch": 1.7430088277054139, + "grad_norm": 0.28860676288604736, + "learning_rate": 9.438746573884938e-07, + "loss": 0.0317, + "step": 56470 + }, + { + "epoch": 1.7433174887338725, + "grad_norm": 0.3121572732925415, + "learning_rate": 9.385527757558909e-07, + "loss": 0.0323, + "step": 56480 + }, + { + "epoch": 1.7436261497623309, + "grad_norm": 0.38573238253593445, + "learning_rate": 9.332457976955644e-07, + "loss": 0.0336, + "step": 56490 + }, + { + "epoch": 1.7439348107907895, + "grad_norm": 0.2921270430088043, + "learning_rate": 9.279537248196247e-07, + "loss": 0.0303, + "step": 56500 + }, + { + "epoch": 1.744243471819248, + "grad_norm": 0.26219409704208374, + "learning_rate": 9.2267655873568e-07, + "loss": 0.0314, + "step": 56510 + }, + { + "epoch": 1.7445521328477067, + "grad_norm": 0.4444236159324646, + "learning_rate": 9.174143010467762e-07, + "loss": 0.036, + "step": 56520 + }, + { + "epoch": 1.7448607938761653, + "grad_norm": 0.5519503355026245, + "learning_rate": 9.12166953351462e-07, + "loss": 0.0339, + "step": 56530 + }, + { + "epoch": 1.7451694549046237, + "grad_norm": 0.30508455634117126, + "learning_rate": 9.069345172437404e-07, + "loss": 0.0367, + "step": 56540 + }, + { + "epoch": 1.7454781159330823, + "grad_norm": 0.30426734685897827, + "learning_rate": 9.017169943130843e-07, + "loss": 0.0338, + "step": 56550 + }, + { + "epoch": 1.7457867769615407, + "grad_norm": 0.36975711584091187, + "learning_rate": 8.965143861444425e-07, + "loss": 0.0298, + "step": 56560 + }, + { + "epoch": 1.7460954379899993, + "grad_norm": 0.38606321811676025, + "learning_rate": 8.91326694318223e-07, + "loss": 0.033, + "step": 56570 + }, + { + "epoch": 1.746404099018458, + "grad_norm": 0.41533002257347107, + "learning_rate": 8.861539204103098e-07, + "loss": 0.0305, + "step": 56580 + }, + { + "epoch": 1.7467127600469166, + "grad_norm": 0.3873472809791565, + "learning_rate": 8.809960659920735e-07, + "loss": 0.0395, + "step": 56590 + }, + { + "epoch": 1.7470214210753752, + "grad_norm": 0.40380850434303284, + "learning_rate": 8.758531326303055e-07, + "loss": 0.03, + "step": 56600 + }, + { + "epoch": 1.7473300821038336, + "grad_norm": 0.3112730383872986, + "learning_rate": 8.707251218873169e-07, + "loss": 0.0311, + "step": 56610 + }, + { + "epoch": 1.7476387431322922, + "grad_norm": 0.34237828850746155, + "learning_rate": 8.656120353208507e-07, + "loss": 0.0338, + "step": 56620 + }, + { + "epoch": 1.7479474041607506, + "grad_norm": 0.31160423159599304, + "learning_rate": 8.605138744841312e-07, + "loss": 0.0341, + "step": 56630 + }, + { + "epoch": 1.7482560651892092, + "grad_norm": 0.31557896733283997, + "learning_rate": 8.554306409258417e-07, + "loss": 0.0299, + "step": 56640 + }, + { + "epoch": 1.7485647262176678, + "grad_norm": 0.3707093298435211, + "learning_rate": 8.503623361901358e-07, + "loss": 0.0304, + "step": 56650 + }, + { + "epoch": 1.7488733872461264, + "grad_norm": 0.4258909821510315, + "learning_rate": 8.453089618166377e-07, + "loss": 0.0336, + "step": 56660 + }, + { + "epoch": 1.749182048274585, + "grad_norm": 0.350748747587204, + "learning_rate": 8.402705193404137e-07, + "loss": 0.0396, + "step": 56670 + }, + { + "epoch": 1.7494907093030434, + "grad_norm": 0.3819507956504822, + "learning_rate": 8.352470102920174e-07, + "loss": 0.0327, + "step": 56680 + }, + { + "epoch": 1.7497993703315018, + "grad_norm": 0.22893407940864563, + "learning_rate": 8.302384361974669e-07, + "loss": 0.0309, + "step": 56690 + }, + { + "epoch": 1.7501080313599604, + "grad_norm": 0.5253183245658875, + "learning_rate": 8.252447985782231e-07, + "loss": 0.0323, + "step": 56700 + }, + { + "epoch": 1.750416692388419, + "grad_norm": 0.29169216752052307, + "learning_rate": 8.202660989512279e-07, + "loss": 0.036, + "step": 56710 + }, + { + "epoch": 1.7507253534168776, + "grad_norm": 0.394599974155426, + "learning_rate": 8.153023388288772e-07, + "loss": 0.033, + "step": 56720 + }, + { + "epoch": 1.7510340144453362, + "grad_norm": 0.29314184188842773, + "learning_rate": 8.103535197190204e-07, + "loss": 0.0339, + "step": 56730 + }, + { + "epoch": 1.7513426754737946, + "grad_norm": 0.38061338663101196, + "learning_rate": 8.054196431249993e-07, + "loss": 0.0327, + "step": 56740 + }, + { + "epoch": 1.7516513365022532, + "grad_norm": 0.3867191672325134, + "learning_rate": 8.005007105455709e-07, + "loss": 0.0323, + "step": 56750 + }, + { + "epoch": 1.7519599975307116, + "grad_norm": 0.35296353697776794, + "learning_rate": 7.95596723474995e-07, + "loss": 0.0342, + "step": 56760 + }, + { + "epoch": 1.7522686585591702, + "grad_norm": 0.3197202682495117, + "learning_rate": 7.907076834029692e-07, + "loss": 0.0332, + "step": 56770 + }, + { + "epoch": 1.7525773195876289, + "grad_norm": 0.20656952261924744, + "learning_rate": 7.858335918146498e-07, + "loss": 0.0316, + "step": 56780 + }, + { + "epoch": 1.7528859806160875, + "grad_norm": 0.4071459174156189, + "learning_rate": 7.809744501906635e-07, + "loss": 0.0295, + "step": 56790 + }, + { + "epoch": 1.753194641644546, + "grad_norm": 0.3132458031177521, + "learning_rate": 7.761302600070797e-07, + "loss": 0.0357, + "step": 56800 + }, + { + "epoch": 1.7535033026730045, + "grad_norm": 0.3397039473056793, + "learning_rate": 7.713010227354545e-07, + "loss": 0.0339, + "step": 56810 + }, + { + "epoch": 1.753811963701463, + "grad_norm": 0.2234487533569336, + "learning_rate": 7.664867398427589e-07, + "loss": 0.0317, + "step": 56820 + }, + { + "epoch": 1.7541206247299215, + "grad_norm": 0.386367529630661, + "learning_rate": 7.616874127914619e-07, + "loss": 0.0323, + "step": 56830 + }, + { + "epoch": 1.75442928575838, + "grad_norm": 0.8971518278121948, + "learning_rate": 7.569030430394641e-07, + "loss": 0.0374, + "step": 56840 + }, + { + "epoch": 1.7547379467868387, + "grad_norm": 0.30268484354019165, + "learning_rate": 7.521336320401306e-07, + "loss": 0.0335, + "step": 56850 + }, + { + "epoch": 1.7550466078152973, + "grad_norm": 0.36130937933921814, + "learning_rate": 7.473791812422915e-07, + "loss": 0.0354, + "step": 56860 + }, + { + "epoch": 1.755355268843756, + "grad_norm": 0.27736252546310425, + "learning_rate": 7.426396920902134e-07, + "loss": 0.035, + "step": 56870 + }, + { + "epoch": 1.7556639298722143, + "grad_norm": 0.2734456956386566, + "learning_rate": 7.379151660236283e-07, + "loss": 0.0351, + "step": 56880 + }, + { + "epoch": 1.755972590900673, + "grad_norm": 0.38918936252593994, + "learning_rate": 7.332056044777324e-07, + "loss": 0.0315, + "step": 56890 + }, + { + "epoch": 1.7562812519291313, + "grad_norm": 0.25528639554977417, + "learning_rate": 7.285110088831537e-07, + "loss": 0.0348, + "step": 56900 + }, + { + "epoch": 1.75658991295759, + "grad_norm": 0.3555104434490204, + "learning_rate": 7.238313806659902e-07, + "loss": 0.0313, + "step": 56910 + }, + { + "epoch": 1.7568985739860485, + "grad_norm": 0.3046354651451111, + "learning_rate": 7.191667212477993e-07, + "loss": 0.031, + "step": 56920 + }, + { + "epoch": 1.7572072350145072, + "grad_norm": 0.3381955623626709, + "learning_rate": 7.145170320455697e-07, + "loss": 0.0326, + "step": 56930 + }, + { + "epoch": 1.7575158960429658, + "grad_norm": 2.252155303955078, + "learning_rate": 7.098823144717604e-07, + "loss": 0.0349, + "step": 56940 + }, + { + "epoch": 1.7578245570714242, + "grad_norm": 0.7058902978897095, + "learning_rate": 7.052625699342674e-07, + "loss": 0.0369, + "step": 56950 + }, + { + "epoch": 1.7581332180998825, + "grad_norm": 0.3479773998260498, + "learning_rate": 7.006577998364628e-07, + "loss": 0.0349, + "step": 56960 + }, + { + "epoch": 1.7584418791283412, + "grad_norm": 0.3944512903690338, + "learning_rate": 6.96068005577133e-07, + "loss": 0.0351, + "step": 56970 + }, + { + "epoch": 1.7587505401567998, + "grad_norm": 0.28638094663619995, + "learning_rate": 6.914931885505627e-07, + "loss": 0.0344, + "step": 56980 + }, + { + "epoch": 1.7590592011852584, + "grad_norm": 0.33968523144721985, + "learning_rate": 6.869333501464347e-07, + "loss": 0.0296, + "step": 56990 + }, + { + "epoch": 1.759367862213717, + "grad_norm": 0.27660253643989563, + "learning_rate": 6.823884917499246e-07, + "loss": 0.0347, + "step": 57000 + }, + { + "epoch": 1.7596765232421756, + "grad_norm": 0.28637203574180603, + "learning_rate": 6.778586147416278e-07, + "loss": 0.0338, + "step": 57010 + }, + { + "epoch": 1.759985184270634, + "grad_norm": 0.4591180980205536, + "learning_rate": 6.733437204976156e-07, + "loss": 0.033, + "step": 57020 + }, + { + "epoch": 1.7602938452990924, + "grad_norm": 0.4799516797065735, + "learning_rate": 6.688438103893857e-07, + "loss": 0.0315, + "step": 57030 + }, + { + "epoch": 1.760602506327551, + "grad_norm": 0.25702664256095886, + "learning_rate": 6.64358885783889e-07, + "loss": 0.0344, + "step": 57040 + }, + { + "epoch": 1.7609111673560096, + "grad_norm": 0.4222659766674042, + "learning_rate": 6.598889480435299e-07, + "loss": 0.0309, + "step": 57050 + }, + { + "epoch": 1.7612198283844682, + "grad_norm": 0.38477593660354614, + "learning_rate": 6.554339985261615e-07, + "loss": 0.0319, + "step": 57060 + }, + { + "epoch": 1.7615284894129268, + "grad_norm": 0.3277072310447693, + "learning_rate": 6.509940385850733e-07, + "loss": 0.0336, + "step": 57070 + }, + { + "epoch": 1.7618371504413852, + "grad_norm": 0.28945136070251465, + "learning_rate": 6.465690695690141e-07, + "loss": 0.0306, + "step": 57080 + }, + { + "epoch": 1.7621458114698438, + "grad_norm": 0.3091687858104706, + "learning_rate": 6.421590928221699e-07, + "loss": 0.0366, + "step": 57090 + }, + { + "epoch": 1.7624544724983022, + "grad_norm": 0.23676130175590515, + "learning_rate": 6.377641096841691e-07, + "loss": 0.0357, + "step": 57100 + }, + { + "epoch": 1.7627631335267608, + "grad_norm": 0.7684627175331116, + "learning_rate": 6.333841214901048e-07, + "loss": 0.0356, + "step": 57110 + }, + { + "epoch": 1.7630717945552195, + "grad_norm": 0.28956711292266846, + "learning_rate": 6.290191295704906e-07, + "loss": 0.0342, + "step": 57120 + }, + { + "epoch": 1.763380455583678, + "grad_norm": 0.2799144983291626, + "learning_rate": 6.246691352513046e-07, + "loss": 0.0333, + "step": 57130 + }, + { + "epoch": 1.7636891166121367, + "grad_norm": 0.33179107308387756, + "learning_rate": 6.203341398539452e-07, + "loss": 0.0306, + "step": 57140 + }, + { + "epoch": 1.763997777640595, + "grad_norm": 0.36287322640419006, + "learning_rate": 6.160141446952872e-07, + "loss": 0.0338, + "step": 57150 + }, + { + "epoch": 1.7643064386690537, + "grad_norm": 0.3393092155456543, + "learning_rate": 6.11709151087625e-07, + "loss": 0.0369, + "step": 57160 + }, + { + "epoch": 1.764615099697512, + "grad_norm": 0.28036051988601685, + "learning_rate": 6.074191603386958e-07, + "loss": 0.0333, + "step": 57170 + }, + { + "epoch": 1.7649237607259707, + "grad_norm": 0.33695077896118164, + "learning_rate": 6.031441737516907e-07, + "loss": 0.0329, + "step": 57180 + }, + { + "epoch": 1.7652324217544293, + "grad_norm": 0.3221655488014221, + "learning_rate": 5.988841926252431e-07, + "loss": 0.0317, + "step": 57190 + }, + { + "epoch": 1.765541082782888, + "grad_norm": 0.2944255471229553, + "learning_rate": 5.946392182534066e-07, + "loss": 0.0339, + "step": 57200 + }, + { + "epoch": 1.7658497438113465, + "grad_norm": 0.4146750271320343, + "learning_rate": 5.90409251925711e-07, + "loss": 0.0347, + "step": 57210 + }, + { + "epoch": 1.766158404839805, + "grad_norm": 0.33428218960762024, + "learning_rate": 5.861942949270949e-07, + "loss": 0.0348, + "step": 57220 + }, + { + "epoch": 1.7664670658682635, + "grad_norm": 0.2970956265926361, + "learning_rate": 5.819943485379564e-07, + "loss": 0.0346, + "step": 57230 + }, + { + "epoch": 1.766775726896722, + "grad_norm": 0.3118849992752075, + "learning_rate": 5.778094140341306e-07, + "loss": 0.0302, + "step": 57240 + }, + { + "epoch": 1.7670843879251805, + "grad_norm": 0.43654653429985046, + "learning_rate": 5.736394926868893e-07, + "loss": 0.0357, + "step": 57250 + }, + { + "epoch": 1.7673930489536391, + "grad_norm": 0.3011344075202942, + "learning_rate": 5.694845857629416e-07, + "loss": 0.0336, + "step": 57260 + }, + { + "epoch": 1.7677017099820977, + "grad_norm": 0.3122650682926178, + "learning_rate": 5.653446945244334e-07, + "loss": 0.0332, + "step": 57270 + }, + { + "epoch": 1.7680103710105564, + "grad_norm": 0.3630503714084625, + "learning_rate": 5.612198202289698e-07, + "loss": 0.0394, + "step": 57280 + }, + { + "epoch": 1.7683190320390147, + "grad_norm": 0.32570669054985046, + "learning_rate": 5.571099641295596e-07, + "loss": 0.0305, + "step": 57290 + }, + { + "epoch": 1.7686276930674731, + "grad_norm": 0.44548463821411133, + "learning_rate": 5.530151274746875e-07, + "loss": 0.0336, + "step": 57300 + }, + { + "epoch": 1.7689363540959318, + "grad_norm": 0.2779359817504883, + "learning_rate": 5.489353115082418e-07, + "loss": 0.036, + "step": 57310 + }, + { + "epoch": 1.7692450151243904, + "grad_norm": 0.4103407561779022, + "learning_rate": 5.4487051746957e-07, + "loss": 0.0361, + "step": 57320 + }, + { + "epoch": 1.769553676152849, + "grad_norm": 0.40970471501350403, + "learning_rate": 5.408207465934511e-07, + "loss": 0.033, + "step": 57330 + }, + { + "epoch": 1.7698623371813076, + "grad_norm": 0.535134494304657, + "learning_rate": 5.3678600011009e-07, + "loss": 0.0346, + "step": 57340 + }, + { + "epoch": 1.7701709982097662, + "grad_norm": 0.25060760974884033, + "learning_rate": 5.327662792451449e-07, + "loss": 0.0371, + "step": 57350 + }, + { + "epoch": 1.7704796592382246, + "grad_norm": 0.28701749444007874, + "learning_rate": 5.287615852196947e-07, + "loss": 0.0408, + "step": 57360 + }, + { + "epoch": 1.770788320266683, + "grad_norm": 0.3151702284812927, + "learning_rate": 5.247719192502665e-07, + "loss": 0.0382, + "step": 57370 + }, + { + "epoch": 1.7710969812951416, + "grad_norm": 0.41918304562568665, + "learning_rate": 5.207972825488128e-07, + "loss": 0.0337, + "step": 57380 + }, + { + "epoch": 1.7714056423236002, + "grad_norm": 0.42279690504074097, + "learning_rate": 5.168376763227178e-07, + "loss": 0.031, + "step": 57390 + }, + { + "epoch": 1.7717143033520588, + "grad_norm": 0.3918517827987671, + "learning_rate": 5.12893101774814e-07, + "loss": 0.0344, + "step": 57400 + }, + { + "epoch": 1.7720229643805174, + "grad_norm": 0.32464924454689026, + "learning_rate": 5.089635601033483e-07, + "loss": 0.0352, + "step": 57410 + }, + { + "epoch": 1.7723316254089758, + "grad_norm": 0.43573805689811707, + "learning_rate": 5.050490525020213e-07, + "loss": 0.0374, + "step": 57420 + }, + { + "epoch": 1.7726402864374344, + "grad_norm": 0.29016298055648804, + "learning_rate": 5.011495801599541e-07, + "loss": 0.0351, + "step": 57430 + }, + { + "epoch": 1.7729489474658928, + "grad_norm": 0.346055805683136, + "learning_rate": 4.972651442616994e-07, + "loss": 0.0294, + "step": 57440 + }, + { + "epoch": 1.7732576084943514, + "grad_norm": 0.2373407930135727, + "learning_rate": 4.933957459872574e-07, + "loss": 0.0325, + "step": 57450 + }, + { + "epoch": 1.77356626952281, + "grad_norm": 0.36258623003959656, + "learning_rate": 4.895413865120324e-07, + "loss": 0.0326, + "step": 57460 + }, + { + "epoch": 1.7738749305512687, + "grad_norm": 0.3273962140083313, + "learning_rate": 4.857020670068935e-07, + "loss": 0.0337, + "step": 57470 + }, + { + "epoch": 1.7741835915797273, + "grad_norm": 0.4046512246131897, + "learning_rate": 4.818777886381132e-07, + "loss": 0.0341, + "step": 57480 + }, + { + "epoch": 1.7744922526081857, + "grad_norm": 0.24349653720855713, + "learning_rate": 4.780685525674122e-07, + "loss": 0.0322, + "step": 57490 + }, + { + "epoch": 1.7748009136366443, + "grad_norm": 0.2829682528972626, + "learning_rate": 4.7427435995193724e-07, + "loss": 0.0311, + "step": 57500 + }, + { + "epoch": 1.7751095746651027, + "grad_norm": 0.2678784132003784, + "learning_rate": 4.7049521194425515e-07, + "loss": 0.0331, + "step": 57510 + }, + { + "epoch": 1.7754182356935613, + "grad_norm": 0.32799333333969116, + "learning_rate": 4.667311096923754e-07, + "loss": 0.0347, + "step": 57520 + }, + { + "epoch": 1.7757268967220199, + "grad_norm": 0.34611913561820984, + "learning_rate": 4.6298205433973895e-07, + "loss": 0.0327, + "step": 57530 + }, + { + "epoch": 1.7760355577504785, + "grad_norm": 0.19358545541763306, + "learning_rate": 4.5924804702520696e-07, + "loss": 0.0306, + "step": 57540 + }, + { + "epoch": 1.7763442187789371, + "grad_norm": 0.4117358326911926, + "learning_rate": 4.5552908888306655e-07, + "loss": 0.0331, + "step": 57550 + }, + { + "epoch": 1.7766528798073955, + "grad_norm": 0.38544929027557373, + "learning_rate": 4.5182518104304185e-07, + "loss": 0.036, + "step": 57560 + }, + { + "epoch": 1.7769615408358541, + "grad_norm": 0.2928087115287781, + "learning_rate": 4.4813632463028277e-07, + "loss": 0.031, + "step": 57570 + }, + { + "epoch": 1.7772702018643125, + "grad_norm": 0.31347838044166565, + "learning_rate": 4.444625207653763e-07, + "loss": 0.0311, + "step": 57580 + }, + { + "epoch": 1.7775788628927711, + "grad_norm": 0.23589091002941132, + "learning_rate": 4.4080377056430753e-07, + "loss": 0.0357, + "step": 57590 + }, + { + "epoch": 1.7778875239212297, + "grad_norm": 0.44620347023010254, + "learning_rate": 4.371600751385263e-07, + "loss": 0.03, + "step": 57600 + }, + { + "epoch": 1.7781961849496883, + "grad_norm": 0.4862222671508789, + "learning_rate": 4.3353143559487495e-07, + "loss": 0.0335, + "step": 57610 + }, + { + "epoch": 1.778504845978147, + "grad_norm": 0.3474585711956024, + "learning_rate": 4.2991785303565513e-07, + "loss": 0.0324, + "step": 57620 + }, + { + "epoch": 1.7788135070066053, + "grad_norm": 0.2807161211967468, + "learning_rate": 4.2631932855856647e-07, + "loss": 0.0324, + "step": 57630 + }, + { + "epoch": 1.7791221680350637, + "grad_norm": 0.30458560585975647, + "learning_rate": 4.2273586325674576e-07, + "loss": 0.0349, + "step": 57640 + }, + { + "epoch": 1.7794308290635223, + "grad_norm": 0.402810275554657, + "learning_rate": 4.1916745821876103e-07, + "loss": 0.0307, + "step": 57650 + }, + { + "epoch": 1.779739490091981, + "grad_norm": 0.3175963759422302, + "learning_rate": 4.156141145285897e-07, + "loss": 0.0325, + "step": 57660 + }, + { + "epoch": 1.7800481511204396, + "grad_norm": 0.36013734340667725, + "learning_rate": 4.1207583326566267e-07, + "loss": 0.0346, + "step": 57670 + }, + { + "epoch": 1.7803568121488982, + "grad_norm": 0.2539544999599457, + "learning_rate": 4.085526155047925e-07, + "loss": 0.0324, + "step": 57680 + }, + { + "epoch": 1.7806654731773566, + "grad_norm": 0.22263766825199127, + "learning_rate": 4.050444623162564e-07, + "loss": 0.0288, + "step": 57690 + }, + { + "epoch": 1.7809741342058152, + "grad_norm": 0.33908411860466003, + "learning_rate": 4.015513747657351e-07, + "loss": 0.0364, + "step": 57700 + }, + { + "epoch": 1.7812827952342736, + "grad_norm": 0.23829835653305054, + "learning_rate": 3.9807335391433554e-07, + "loss": 0.0299, + "step": 57710 + }, + { + "epoch": 1.7815914562627322, + "grad_norm": 0.3502861559391022, + "learning_rate": 3.946104008185847e-07, + "loss": 0.0341, + "step": 57720 + }, + { + "epoch": 1.7819001172911908, + "grad_norm": 0.27886053919792175, + "learning_rate": 3.9116251653044113e-07, + "loss": 0.0336, + "step": 57730 + }, + { + "epoch": 1.7822087783196494, + "grad_norm": 0.38831695914268494, + "learning_rate": 3.877297020972781e-07, + "loss": 0.0352, + "step": 57740 + }, + { + "epoch": 1.782517439348108, + "grad_norm": 0.3834664225578308, + "learning_rate": 3.8431195856190036e-07, + "loss": 0.0327, + "step": 57750 + }, + { + "epoch": 1.7828261003765664, + "grad_norm": 0.3779309391975403, + "learning_rate": 3.8090928696251635e-07, + "loss": 0.033, + "step": 57760 + }, + { + "epoch": 1.783134761405025, + "grad_norm": 0.33471962809562683, + "learning_rate": 3.77521688332777e-07, + "loss": 0.0338, + "step": 57770 + }, + { + "epoch": 1.7834434224334834, + "grad_norm": 0.26029881834983826, + "learning_rate": 3.741491637017425e-07, + "loss": 0.031, + "step": 57780 + }, + { + "epoch": 1.783752083461942, + "grad_norm": 0.2696375548839569, + "learning_rate": 3.707917140939043e-07, + "loss": 0.0353, + "step": 57790 + }, + { + "epoch": 1.7840607444904006, + "grad_norm": 0.7434337139129639, + "learning_rate": 3.6744934052915235e-07, + "loss": 0.0362, + "step": 57800 + }, + { + "epoch": 1.7843694055188593, + "grad_norm": 0.27561283111572266, + "learning_rate": 3.641220440228188e-07, + "loss": 0.0334, + "step": 57810 + }, + { + "epoch": 1.7846780665473179, + "grad_norm": 0.2814045250415802, + "learning_rate": 3.608098255856562e-07, + "loss": 0.028, + "step": 57820 + }, + { + "epoch": 1.7849867275757763, + "grad_norm": 0.43785691261291504, + "learning_rate": 3.575126862238154e-07, + "loss": 0.0342, + "step": 57830 + }, + { + "epoch": 1.7852953886042349, + "grad_norm": 0.31822526454925537, + "learning_rate": 3.5423062693888955e-07, + "loss": 0.0324, + "step": 57840 + }, + { + "epoch": 1.7856040496326933, + "grad_norm": 0.33898523449897766, + "learning_rate": 3.509636487278756e-07, + "loss": 0.0327, + "step": 57850 + }, + { + "epoch": 1.7859127106611519, + "grad_norm": 0.3086783289909363, + "learning_rate": 3.4771175258320186e-07, + "loss": 0.032, + "step": 57860 + }, + { + "epoch": 1.7862213716896105, + "grad_norm": 0.4849085807800293, + "learning_rate": 3.4447493949270047e-07, + "loss": 0.029, + "step": 57870 + }, + { + "epoch": 1.786530032718069, + "grad_norm": 0.38700366020202637, + "learning_rate": 3.4125321043964045e-07, + "loss": 0.0297, + "step": 57880 + }, + { + "epoch": 1.7868386937465277, + "grad_norm": 0.39977219700813293, + "learning_rate": 3.380465664026833e-07, + "loss": 0.0363, + "step": 57890 + }, + { + "epoch": 1.787147354774986, + "grad_norm": 0.43515223264694214, + "learning_rate": 3.348550083559388e-07, + "loss": 0.0357, + "step": 57900 + }, + { + "epoch": 1.7874560158034447, + "grad_norm": 0.31023526191711426, + "learning_rate": 3.316785372689091e-07, + "loss": 0.0341, + "step": 57910 + }, + { + "epoch": 1.787764676831903, + "grad_norm": 0.272892028093338, + "learning_rate": 3.285171541065224e-07, + "loss": 0.0345, + "step": 57920 + }, + { + "epoch": 1.7880733378603617, + "grad_norm": 0.27935364842414856, + "learning_rate": 3.253708598291272e-07, + "loss": 0.0335, + "step": 57930 + }, + { + "epoch": 1.7883819988888203, + "grad_norm": 0.358064740896225, + "learning_rate": 3.2223965539248116e-07, + "loss": 0.0324, + "step": 57940 + }, + { + "epoch": 1.788690659917279, + "grad_norm": 0.26673877239227295, + "learning_rate": 3.1912354174776227e-07, + "loss": 0.0324, + "step": 57950 + }, + { + "epoch": 1.7889993209457375, + "grad_norm": 0.36019301414489746, + "learning_rate": 3.1602251984155786e-07, + "loss": 0.033, + "step": 57960 + }, + { + "epoch": 1.789307981974196, + "grad_norm": 0.4837923049926758, + "learning_rate": 3.1293659061589207e-07, + "loss": 0.0334, + "step": 57970 + }, + { + "epoch": 1.7896166430026543, + "grad_norm": 0.28714656829833984, + "learning_rate": 3.098657550081707e-07, + "loss": 0.0305, + "step": 57980 + }, + { + "epoch": 1.789925304031113, + "grad_norm": 0.45382800698280334, + "learning_rate": 3.068100139512475e-07, + "loss": 0.0338, + "step": 57990 + }, + { + "epoch": 1.7902339650595716, + "grad_norm": 0.3357143998146057, + "learning_rate": 3.037693683733689e-07, + "loss": 0.033, + "step": 58000 + }, + { + "epoch": 1.7905426260880302, + "grad_norm": 0.35550183057785034, + "learning_rate": 3.007438191982015e-07, + "loss": 0.0319, + "step": 58010 + }, + { + "epoch": 1.7908512871164888, + "grad_norm": 0.37523335218429565, + "learning_rate": 2.9773336734482684e-07, + "loss": 0.0343, + "step": 58020 + }, + { + "epoch": 1.7911599481449472, + "grad_norm": 0.26290538907051086, + "learning_rate": 2.9473801372774667e-07, + "loss": 0.0319, + "step": 58030 + }, + { + "epoch": 1.7914686091734058, + "grad_norm": 0.3056303858757019, + "learning_rate": 2.91757759256861e-07, + "loss": 0.0354, + "step": 58040 + }, + { + "epoch": 1.7917772702018642, + "grad_norm": 0.31242650747299194, + "learning_rate": 2.887926048375067e-07, + "loss": 0.032, + "step": 58050 + }, + { + "epoch": 1.7920859312303228, + "grad_norm": 0.3653290867805481, + "learning_rate": 2.858425513704022e-07, + "loss": 0.0339, + "step": 58060 + }, + { + "epoch": 1.7923945922587814, + "grad_norm": 0.3132387697696686, + "learning_rate": 2.8290759975170834e-07, + "loss": 0.0378, + "step": 58070 + }, + { + "epoch": 1.79270325328724, + "grad_norm": 0.3127036392688751, + "learning_rate": 2.799877508729787e-07, + "loss": 0.0315, + "step": 58080 + }, + { + "epoch": 1.7930119143156986, + "grad_norm": 0.2688280940055847, + "learning_rate": 2.770830056211926e-07, + "loss": 0.0345, + "step": 58090 + }, + { + "epoch": 1.793320575344157, + "grad_norm": 1.0642770528793335, + "learning_rate": 2.741933648787331e-07, + "loss": 0.0325, + "step": 58100 + }, + { + "epoch": 1.7936292363726156, + "grad_norm": 0.2807531952857971, + "learning_rate": 2.7131882952339263e-07, + "loss": 0.0311, + "step": 58110 + }, + { + "epoch": 1.793937897401074, + "grad_norm": 0.39542341232299805, + "learning_rate": 2.684594004283836e-07, + "loss": 0.0336, + "step": 58120 + }, + { + "epoch": 1.7942465584295326, + "grad_norm": 0.3332241475582123, + "learning_rate": 2.6561507846232234e-07, + "loss": 0.0338, + "step": 58130 + }, + { + "epoch": 1.7945552194579912, + "grad_norm": 0.28782719373703003, + "learning_rate": 2.6278586448924005e-07, + "loss": 0.0326, + "step": 58140 + }, + { + "epoch": 1.7948638804864498, + "grad_norm": 0.6627581119537354, + "learning_rate": 2.5997175936857685e-07, + "loss": 0.0299, + "step": 58150 + }, + { + "epoch": 1.7951725415149085, + "grad_norm": 0.327847957611084, + "learning_rate": 2.57172763955188e-07, + "loss": 0.0359, + "step": 58160 + }, + { + "epoch": 1.7954812025433668, + "grad_norm": 0.30479639768600464, + "learning_rate": 2.543888790993265e-07, + "loss": 0.029, + "step": 58170 + }, + { + "epoch": 1.7957898635718255, + "grad_norm": 0.26315104961395264, + "learning_rate": 2.5162010564666607e-07, + "loss": 0.0331, + "step": 58180 + }, + { + "epoch": 1.7960985246002839, + "grad_norm": 0.39230120182037354, + "learning_rate": 2.488664444382893e-07, + "loss": 0.0358, + "step": 58190 + }, + { + "epoch": 1.7964071856287425, + "grad_norm": 0.44802138209342957, + "learning_rate": 2.461278963106828e-07, + "loss": 0.0353, + "step": 58200 + }, + { + "epoch": 1.796715846657201, + "grad_norm": 0.4937394857406616, + "learning_rate": 2.434044620957421e-07, + "loss": 0.0358, + "step": 58210 + }, + { + "epoch": 1.7970245076856597, + "grad_norm": 0.4664301574230194, + "learning_rate": 2.406961426207832e-07, + "loss": 0.0356, + "step": 58220 + }, + { + "epoch": 1.7973331687141183, + "grad_norm": 0.31520652770996094, + "learning_rate": 2.380029387085203e-07, + "loss": 0.0335, + "step": 58230 + }, + { + "epoch": 1.7976418297425767, + "grad_norm": 0.3143216371536255, + "learning_rate": 2.353248511770767e-07, + "loss": 0.033, + "step": 58240 + }, + { + "epoch": 1.7979504907710353, + "grad_norm": 0.3063029944896698, + "learning_rate": 2.3266188083997942e-07, + "loss": 0.0326, + "step": 58250 + }, + { + "epoch": 1.7982591517994937, + "grad_norm": 0.467177152633667, + "learning_rate": 2.3001402850617027e-07, + "loss": 0.0333, + "step": 58260 + }, + { + "epoch": 1.7985678128279523, + "grad_norm": 0.25785523653030396, + "learning_rate": 2.2738129498000581e-07, + "loss": 0.031, + "step": 58270 + }, + { + "epoch": 1.798876473856411, + "grad_norm": 0.3622733950614929, + "learning_rate": 2.2476368106122414e-07, + "loss": 0.0297, + "step": 58280 + }, + { + "epoch": 1.7991851348848695, + "grad_norm": 0.2540142834186554, + "learning_rate": 2.2216118754500582e-07, + "loss": 0.0336, + "step": 58290 + }, + { + "epoch": 1.7994937959133281, + "grad_norm": 0.3117138147354126, + "learning_rate": 2.1957381522190735e-07, + "loss": 0.0373, + "step": 58300 + }, + { + "epoch": 1.7998024569417865, + "grad_norm": 0.28259703516960144, + "learning_rate": 2.1700156487790558e-07, + "loss": 0.0344, + "step": 58310 + }, + { + "epoch": 1.800111117970245, + "grad_norm": 1.8455374240875244, + "learning_rate": 2.1444443729439213e-07, + "loss": 0.0331, + "step": 58320 + }, + { + "epoch": 1.8004197789987035, + "grad_norm": 0.3057728111743927, + "learning_rate": 2.1190243324814007e-07, + "loss": 0.0353, + "step": 58330 + }, + { + "epoch": 1.8007284400271621, + "grad_norm": 0.40700405836105347, + "learning_rate": 2.0937555351135395e-07, + "loss": 0.0345, + "step": 58340 + }, + { + "epoch": 1.8010371010556208, + "grad_norm": 0.35700759291648865, + "learning_rate": 2.0686379885162532e-07, + "loss": 0.0338, + "step": 58350 + }, + { + "epoch": 1.8013457620840794, + "grad_norm": 0.5757080316543579, + "learning_rate": 2.0436717003196604e-07, + "loss": 0.0332, + "step": 58360 + }, + { + "epoch": 1.8016544231125378, + "grad_norm": 0.359232097864151, + "learning_rate": 2.0188566781078057e-07, + "loss": 0.0359, + "step": 58370 + }, + { + "epoch": 1.8019630841409964, + "grad_norm": 0.23079843819141388, + "learning_rate": 1.994192929418881e-07, + "loss": 0.0328, + "step": 58380 + }, + { + "epoch": 1.8022717451694548, + "grad_norm": 0.27538007497787476, + "learning_rate": 1.9696804617451158e-07, + "loss": 0.0327, + "step": 58390 + }, + { + "epoch": 1.8025804061979134, + "grad_norm": 0.6477785110473633, + "learning_rate": 1.9453192825326093e-07, + "loss": 0.0324, + "step": 58400 + }, + { + "epoch": 1.802889067226372, + "grad_norm": 0.2742314040660858, + "learning_rate": 1.9211093991817753e-07, + "loss": 0.0311, + "step": 58410 + }, + { + "epoch": 1.8031977282548306, + "grad_norm": 0.5331084132194519, + "learning_rate": 1.8970508190468973e-07, + "loss": 0.0319, + "step": 58420 + }, + { + "epoch": 1.8035063892832892, + "grad_norm": 0.3425188958644867, + "learning_rate": 1.8731435494362958e-07, + "loss": 0.0355, + "step": 58430 + }, + { + "epoch": 1.8038150503117476, + "grad_norm": 0.3236185610294342, + "learning_rate": 1.849387597612495e-07, + "loss": 0.0307, + "step": 58440 + }, + { + "epoch": 1.8041237113402062, + "grad_norm": 0.6415771245956421, + "learning_rate": 1.8257829707917228e-07, + "loss": 0.0336, + "step": 58450 + }, + { + "epoch": 1.8044323723686646, + "grad_norm": 0.3526233434677124, + "learning_rate": 1.8023296761446317e-07, + "loss": 0.0383, + "step": 58460 + }, + { + "epoch": 1.8047410333971232, + "grad_norm": 0.3052462935447693, + "learning_rate": 1.7790277207956341e-07, + "loss": 0.0343, + "step": 58470 + }, + { + "epoch": 1.8050496944255818, + "grad_norm": 0.37055331468582153, + "learning_rate": 1.7558771118232343e-07, + "loss": 0.0323, + "step": 58480 + }, + { + "epoch": 1.8053583554540404, + "grad_norm": 0.3832804262638092, + "learning_rate": 1.7328778562599734e-07, + "loss": 0.0312, + "step": 58490 + }, + { + "epoch": 1.805667016482499, + "grad_norm": 0.379048228263855, + "learning_rate": 1.7100299610924298e-07, + "loss": 0.0348, + "step": 58500 + }, + { + "epoch": 1.8059756775109574, + "grad_norm": 0.3235965669155121, + "learning_rate": 1.6873334332612733e-07, + "loss": 0.034, + "step": 58510 + }, + { + "epoch": 1.806284338539416, + "grad_norm": 0.464714914560318, + "learning_rate": 1.6647882796609894e-07, + "loss": 0.0318, + "step": 58520 + }, + { + "epoch": 1.8065929995678744, + "grad_norm": 0.3035613000392914, + "learning_rate": 1.6423945071402102e-07, + "loss": 0.0354, + "step": 58530 + }, + { + "epoch": 1.806901660596333, + "grad_norm": 0.2878853678703308, + "learning_rate": 1.6201521225016614e-07, + "loss": 0.0352, + "step": 58540 + }, + { + "epoch": 1.8072103216247917, + "grad_norm": 0.42007893323898315, + "learning_rate": 1.598061132501938e-07, + "loss": 0.0341, + "step": 58550 + }, + { + "epoch": 1.8075189826532503, + "grad_norm": 0.33289414644241333, + "learning_rate": 1.576121543851672e-07, + "loss": 0.0311, + "step": 58560 + }, + { + "epoch": 1.807827643681709, + "grad_norm": 0.29431864619255066, + "learning_rate": 1.5543333632155876e-07, + "loss": 0.0355, + "step": 58570 + }, + { + "epoch": 1.8081363047101673, + "grad_norm": 0.35699236392974854, + "learning_rate": 1.5326965972123352e-07, + "loss": 0.0363, + "step": 58580 + }, + { + "epoch": 1.8084449657386257, + "grad_norm": 0.30631351470947266, + "learning_rate": 1.5112112524146016e-07, + "loss": 0.0341, + "step": 58590 + }, + { + "epoch": 1.8087536267670843, + "grad_norm": 0.3578760325908661, + "learning_rate": 1.4898773353489992e-07, + "loss": 0.0341, + "step": 58600 + }, + { + "epoch": 1.809062287795543, + "grad_norm": 0.2977222800254822, + "learning_rate": 1.4686948524962884e-07, + "loss": 0.0359, + "step": 58610 + }, + { + "epoch": 1.8093709488240015, + "grad_norm": 0.2622911334037781, + "learning_rate": 1.4476638102911556e-07, + "loss": 0.0301, + "step": 58620 + }, + { + "epoch": 1.8096796098524601, + "grad_norm": 0.34874632954597473, + "learning_rate": 1.4267842151222123e-07, + "loss": 0.0329, + "step": 58630 + }, + { + "epoch": 1.8099882708809187, + "grad_norm": 0.36763474345207214, + "learning_rate": 1.4060560733321626e-07, + "loss": 0.03, + "step": 58640 + }, + { + "epoch": 1.8102969319093771, + "grad_norm": 0.3023255169391632, + "learning_rate": 1.385479391217692e-07, + "loss": 0.0301, + "step": 58650 + }, + { + "epoch": 1.8106055929378355, + "grad_norm": 0.31529340147972107, + "learning_rate": 1.365054175029412e-07, + "loss": 0.0372, + "step": 58660 + }, + { + "epoch": 1.8109142539662941, + "grad_norm": 0.36284148693084717, + "learning_rate": 1.3447804309719702e-07, + "loss": 0.0371, + "step": 58670 + }, + { + "epoch": 1.8112229149947527, + "grad_norm": 0.29496118426322937, + "learning_rate": 1.3246581652040512e-07, + "loss": 0.0386, + "step": 58680 + }, + { + "epoch": 1.8115315760232114, + "grad_norm": 0.27546387910842896, + "learning_rate": 1.3046873838381546e-07, + "loss": 0.0353, + "step": 58690 + }, + { + "epoch": 1.81184023705167, + "grad_norm": 0.32157865166664124, + "learning_rate": 1.2848680929409828e-07, + "loss": 0.0327, + "step": 58700 + }, + { + "epoch": 1.8121488980801284, + "grad_norm": 0.3320658802986145, + "learning_rate": 1.2652002985331091e-07, + "loss": 0.0335, + "step": 58710 + }, + { + "epoch": 1.812457559108587, + "grad_norm": 0.34203651547431946, + "learning_rate": 1.2456840065889764e-07, + "loss": 0.0334, + "step": 58720 + }, + { + "epoch": 1.8127662201370454, + "grad_norm": 0.40443214774131775, + "learning_rate": 1.226319223037231e-07, + "loss": 0.0308, + "step": 58730 + }, + { + "epoch": 1.813074881165504, + "grad_norm": 0.2835387885570526, + "learning_rate": 1.2071059537603902e-07, + "loss": 0.0342, + "step": 58740 + }, + { + "epoch": 1.8133835421939626, + "grad_norm": 0.4553341567516327, + "learning_rate": 1.1880442045948403e-07, + "loss": 0.0321, + "step": 58750 + }, + { + "epoch": 1.8136922032224212, + "grad_norm": 0.31143778562545776, + "learning_rate": 1.1691339813311164e-07, + "loss": 0.0349, + "step": 58760 + }, + { + "epoch": 1.8140008642508798, + "grad_norm": 0.29210636019706726, + "learning_rate": 1.1503752897136233e-07, + "loss": 0.034, + "step": 58770 + }, + { + "epoch": 1.8143095252793382, + "grad_norm": 0.3351101875305176, + "learning_rate": 1.1317681354407472e-07, + "loss": 0.0359, + "step": 58780 + }, + { + "epoch": 1.8146181863077968, + "grad_norm": 0.4238208830356598, + "learning_rate": 1.1133125241649111e-07, + "loss": 0.0313, + "step": 58790 + }, + { + "epoch": 1.8149268473362552, + "grad_norm": 0.2916317880153656, + "learning_rate": 1.0950084614922973e-07, + "loss": 0.0403, + "step": 58800 + }, + { + "epoch": 1.8152355083647138, + "grad_norm": 0.28415387868881226, + "learning_rate": 1.0768559529834021e-07, + "loss": 0.0366, + "step": 58810 + }, + { + "epoch": 1.8155441693931724, + "grad_norm": 0.45983338356018066, + "learning_rate": 1.0588550041522594e-07, + "loss": 0.0332, + "step": 58820 + }, + { + "epoch": 1.815852830421631, + "grad_norm": 0.4245419204235077, + "learning_rate": 1.0410056204672169e-07, + "loss": 0.0319, + "step": 58830 + }, + { + "epoch": 1.8161614914500896, + "grad_norm": 0.3489139974117279, + "learning_rate": 1.0233078073504376e-07, + "loss": 0.0309, + "step": 58840 + }, + { + "epoch": 1.816470152478548, + "grad_norm": 0.35845285654067993, + "learning_rate": 1.0057615701780654e-07, + "loss": 0.0335, + "step": 58850 + }, + { + "epoch": 1.8167788135070067, + "grad_norm": 0.3228519856929779, + "learning_rate": 9.883669142801144e-08, + "loss": 0.0304, + "step": 58860 + }, + { + "epoch": 1.817087474535465, + "grad_norm": 0.23596403002738953, + "learning_rate": 9.711238449406356e-08, + "loss": 0.0297, + "step": 58870 + }, + { + "epoch": 1.8173961355639237, + "grad_norm": 0.3447478413581848, + "learning_rate": 9.540323673976614e-08, + "loss": 0.0332, + "step": 58880 + }, + { + "epoch": 1.8177047965923823, + "grad_norm": 0.5666099190711975, + "learning_rate": 9.370924868430942e-08, + "loss": 0.0361, + "step": 58890 + }, + { + "epoch": 1.8180134576208409, + "grad_norm": 0.2590516209602356, + "learning_rate": 9.203042084228175e-08, + "loss": 0.0328, + "step": 58900 + }, + { + "epoch": 1.8183221186492995, + "grad_norm": 0.27993735671043396, + "learning_rate": 9.036675372366965e-08, + "loss": 0.0333, + "step": 58910 + }, + { + "epoch": 1.8186307796777579, + "grad_norm": 0.21783682703971863, + "learning_rate": 8.871824783385218e-08, + "loss": 0.0331, + "step": 58920 + }, + { + "epoch": 1.8189394407062163, + "grad_norm": 0.29595258831977844, + "learning_rate": 8.7084903673601e-08, + "loss": 0.0368, + "step": 58930 + }, + { + "epoch": 1.8192481017346749, + "grad_norm": 0.2358887791633606, + "learning_rate": 8.546672173908032e-08, + "loss": 0.0344, + "step": 58940 + }, + { + "epoch": 1.8195567627631335, + "grad_norm": 0.42878493666648865, + "learning_rate": 8.386370252185249e-08, + "loss": 0.0375, + "step": 58950 + }, + { + "epoch": 1.819865423791592, + "grad_norm": 0.3177928626537323, + "learning_rate": 8.227584650887243e-08, + "loss": 0.0315, + "step": 58960 + }, + { + "epoch": 1.8201740848200507, + "grad_norm": 0.3173290193080902, + "learning_rate": 8.070315418249319e-08, + "loss": 0.0327, + "step": 58970 + }, + { + "epoch": 1.8204827458485093, + "grad_norm": 0.26592254638671875, + "learning_rate": 7.914562602044929e-08, + "loss": 0.0316, + "step": 58980 + }, + { + "epoch": 1.8207914068769677, + "grad_norm": 0.35370391607284546, + "learning_rate": 7.76032624958789e-08, + "loss": 0.0339, + "step": 58990 + }, + { + "epoch": 1.8211000679054261, + "grad_norm": 0.27200648188591003, + "learning_rate": 7.607606407731282e-08, + "loss": 0.0281, + "step": 59000 + }, + { + "epoch": 1.8214087289338847, + "grad_norm": 0.28224125504493713, + "learning_rate": 7.45640312286744e-08, + "loss": 0.0319, + "step": 59010 + }, + { + "epoch": 1.8217173899623433, + "grad_norm": 0.28347307443618774, + "learning_rate": 7.306716440927952e-08, + "loss": 0.0321, + "step": 59020 + }, + { + "epoch": 1.822026050990802, + "grad_norm": 0.2546809911727905, + "learning_rate": 7.15854640738367e-08, + "loss": 0.0316, + "step": 59030 + }, + { + "epoch": 1.8223347120192606, + "grad_norm": 0.2528218924999237, + "learning_rate": 7.011893067244701e-08, + "loss": 0.0408, + "step": 59040 + }, + { + "epoch": 1.822643373047719, + "grad_norm": 0.39743250608444214, + "learning_rate": 6.866756465060408e-08, + "loss": 0.0343, + "step": 59050 + }, + { + "epoch": 1.8229520340761776, + "grad_norm": 0.2719951272010803, + "learning_rate": 6.723136644918859e-08, + "loss": 0.0365, + "step": 59060 + }, + { + "epoch": 1.823260695104636, + "grad_norm": 0.27828192710876465, + "learning_rate": 6.581033650449042e-08, + "loss": 0.0345, + "step": 59070 + }, + { + "epoch": 1.8235693561330946, + "grad_norm": 0.3257664442062378, + "learning_rate": 6.440447524817539e-08, + "loss": 0.0373, + "step": 59080 + }, + { + "epoch": 1.8238780171615532, + "grad_norm": 0.3502618670463562, + "learning_rate": 6.301378310730743e-08, + "loss": 0.0356, + "step": 59090 + }, + { + "epoch": 1.8241866781900118, + "grad_norm": 0.3330199420452118, + "learning_rate": 6.163826050434307e-08, + "loss": 0.0318, + "step": 59100 + }, + { + "epoch": 1.8244953392184704, + "grad_norm": 0.4264217019081116, + "learning_rate": 6.027790785713139e-08, + "loss": 0.0284, + "step": 59110 + }, + { + "epoch": 1.8248040002469288, + "grad_norm": 0.6285958886146545, + "learning_rate": 5.89327255789085e-08, + "loss": 0.0295, + "step": 59120 + }, + { + "epoch": 1.8251126612753874, + "grad_norm": 0.3455996513366699, + "learning_rate": 5.7602714078303085e-08, + "loss": 0.0327, + "step": 59130 + }, + { + "epoch": 1.8254213223038458, + "grad_norm": 0.30604392290115356, + "learning_rate": 5.628787375934197e-08, + "loss": 0.0334, + "step": 59140 + }, + { + "epoch": 1.8257299833323044, + "grad_norm": 0.4150135815143585, + "learning_rate": 5.498820502143898e-08, + "loss": 0.0321, + "step": 59150 + }, + { + "epoch": 1.826038644360763, + "grad_norm": 0.5100867748260498, + "learning_rate": 5.370370825939497e-08, + "loss": 0.036, + "step": 59160 + }, + { + "epoch": 1.8263473053892216, + "grad_norm": 0.3067622780799866, + "learning_rate": 5.243438386340893e-08, + "loss": 0.0368, + "step": 59170 + }, + { + "epoch": 1.8266559664176802, + "grad_norm": 0.3017074763774872, + "learning_rate": 5.118023221907242e-08, + "loss": 0.0341, + "step": 59180 + }, + { + "epoch": 1.8269646274461386, + "grad_norm": 0.38297221064567566, + "learning_rate": 4.994125370735292e-08, + "loss": 0.0302, + "step": 59190 + }, + { + "epoch": 1.8272732884745972, + "grad_norm": 0.5319650769233704, + "learning_rate": 4.871744870462713e-08, + "loss": 0.0393, + "step": 59200 + }, + { + "epoch": 1.8275819495030556, + "grad_norm": 0.3735000491142273, + "learning_rate": 4.7508817582658794e-08, + "loss": 0.0336, + "step": 59210 + }, + { + "epoch": 1.8278906105315142, + "grad_norm": 0.42647993564605713, + "learning_rate": 4.631536070858755e-08, + "loss": 0.0326, + "step": 59220 + }, + { + "epoch": 1.8281992715599729, + "grad_norm": 0.23314014077186584, + "learning_rate": 4.513707844495674e-08, + "loss": 0.0315, + "step": 59230 + }, + { + "epoch": 1.8285079325884315, + "grad_norm": 0.27521321177482605, + "learning_rate": 4.3973971149702255e-08, + "loss": 0.0307, + "step": 59240 + }, + { + "epoch": 1.82881659361689, + "grad_norm": 0.409643292427063, + "learning_rate": 4.2826039176147025e-08, + "loss": 0.0338, + "step": 59250 + }, + { + "epoch": 1.8291252546453485, + "grad_norm": 0.5842083096504211, + "learning_rate": 4.169328287299545e-08, + "loss": 0.0308, + "step": 59260 + }, + { + "epoch": 1.8294339156738069, + "grad_norm": 0.24659791588783264, + "learning_rate": 4.057570258435006e-08, + "loss": 0.0319, + "step": 59270 + }, + { + "epoch": 1.8297425767022655, + "grad_norm": 0.26821210980415344, + "learning_rate": 3.947329864970595e-08, + "loss": 0.0304, + "step": 59280 + }, + { + "epoch": 1.830051237730724, + "grad_norm": 0.24326805770397186, + "learning_rate": 3.8386071403939686e-08, + "loss": 0.0382, + "step": 59290 + }, + { + "epoch": 1.8303598987591827, + "grad_norm": 0.300235390663147, + "learning_rate": 3.731402117733152e-08, + "loss": 0.0276, + "step": 59300 + }, + { + "epoch": 1.8306685597876413, + "grad_norm": 0.3535788059234619, + "learning_rate": 3.625714829552651e-08, + "loss": 0.0313, + "step": 59310 + }, + { + "epoch": 1.8309772208160997, + "grad_norm": 0.3577711284160614, + "learning_rate": 3.5215453079590065e-08, + "loss": 0.0331, + "step": 59320 + }, + { + "epoch": 1.8312858818445583, + "grad_norm": 0.3168996572494507, + "learning_rate": 3.4188935845952396e-08, + "loss": 0.034, + "step": 59330 + }, + { + "epoch": 1.8315945428730167, + "grad_norm": 0.2826961576938629, + "learning_rate": 3.3177596906447396e-08, + "loss": 0.0307, + "step": 59340 + }, + { + "epoch": 1.8319032039014753, + "grad_norm": 0.2880949079990387, + "learning_rate": 3.218143656829043e-08, + "loss": 0.0345, + "step": 59350 + }, + { + "epoch": 1.832211864929934, + "grad_norm": 0.3787905275821686, + "learning_rate": 3.120045513408387e-08, + "loss": 0.0329, + "step": 59360 + }, + { + "epoch": 1.8325205259583925, + "grad_norm": 0.28687694668769836, + "learning_rate": 3.02346529018338e-08, + "loss": 0.0334, + "step": 59370 + }, + { + "epoch": 1.8328291869868512, + "grad_norm": 0.2719404697418213, + "learning_rate": 2.9284030164922204e-08, + "loss": 0.0309, + "step": 59380 + }, + { + "epoch": 1.8331378480153095, + "grad_norm": 0.21338437497615814, + "learning_rate": 2.8348587212123634e-08, + "loss": 0.0319, + "step": 59390 + }, + { + "epoch": 1.8334465090437682, + "grad_norm": 0.28703340888023376, + "learning_rate": 2.7428324327594125e-08, + "loss": 0.0347, + "step": 59400 + }, + { + "epoch": 1.8337551700722265, + "grad_norm": 0.26183703541755676, + "learning_rate": 2.6523241790893383e-08, + "loss": 0.0381, + "step": 59410 + }, + { + "epoch": 1.8340638311006852, + "grad_norm": 0.9841156005859375, + "learning_rate": 2.563333987695704e-08, + "loss": 0.036, + "step": 59420 + }, + { + "epoch": 1.8343724921291438, + "grad_norm": 0.34447720646858215, + "learning_rate": 2.4758618856118852e-08, + "loss": 0.0297, + "step": 59430 + }, + { + "epoch": 1.8346811531576024, + "grad_norm": 0.302499920129776, + "learning_rate": 2.3899078994088497e-08, + "loss": 0.0334, + "step": 59440 + }, + { + "epoch": 1.834989814186061, + "grad_norm": 0.376038134098053, + "learning_rate": 2.3054720551973775e-08, + "loss": 0.0336, + "step": 59450 + }, + { + "epoch": 1.8352984752145194, + "grad_norm": 0.3811785876750946, + "learning_rate": 2.222554378627506e-08, + "loss": 0.0336, + "step": 59460 + }, + { + "epoch": 1.835607136242978, + "grad_norm": 0.2618066072463989, + "learning_rate": 2.1411548948868653e-08, + "loss": 0.0366, + "step": 59470 + }, + { + "epoch": 1.8359157972714364, + "grad_norm": 0.4041125178337097, + "learning_rate": 2.0612736287023426e-08, + "loss": 0.036, + "step": 59480 + }, + { + "epoch": 1.836224458299895, + "grad_norm": 0.4843859374523163, + "learning_rate": 1.9829106043400826e-08, + "loss": 0.0321, + "step": 59490 + }, + { + "epoch": 1.8365331193283536, + "grad_norm": 0.3168388605117798, + "learning_rate": 1.9060658456043768e-08, + "loss": 0.0372, + "step": 59500 + }, + { + "epoch": 1.8368417803568122, + "grad_norm": 0.22596968710422516, + "learning_rate": 1.830739375838775e-08, + "loss": 0.0341, + "step": 59510 + }, + { + "epoch": 1.8371504413852708, + "grad_norm": 0.2840762734413147, + "learning_rate": 1.7569312179260832e-08, + "loss": 0.0334, + "step": 59520 + }, + { + "epoch": 1.8374591024137292, + "grad_norm": 0.21939513087272644, + "learning_rate": 1.684641394286146e-08, + "loss": 0.0374, + "step": 59530 + }, + { + "epoch": 1.8377677634421878, + "grad_norm": 0.29637250304222107, + "learning_rate": 1.6138699268797296e-08, + "loss": 0.0329, + "step": 59540 + }, + { + "epoch": 1.8380764244706462, + "grad_norm": 0.3733431398868561, + "learning_rate": 1.5446168372046376e-08, + "loss": 0.0365, + "step": 59550 + }, + { + "epoch": 1.8383850854991048, + "grad_norm": 0.2610968053340912, + "learning_rate": 1.4768821462984861e-08, + "loss": 0.0337, + "step": 59560 + }, + { + "epoch": 1.8386937465275635, + "grad_norm": 1.013060212135315, + "learning_rate": 1.4106658747370383e-08, + "loss": 0.0365, + "step": 59570 + }, + { + "epoch": 1.839002407556022, + "grad_norm": 0.37035319209098816, + "learning_rate": 1.3459680426353149e-08, + "loss": 0.0304, + "step": 59580 + }, + { + "epoch": 1.8393110685844807, + "grad_norm": 0.5135836601257324, + "learning_rate": 1.2827886696464841e-08, + "loss": 0.037, + "step": 59590 + }, + { + "epoch": 1.839619729612939, + "grad_norm": 0.20988604426383972, + "learning_rate": 1.2211277749635264e-08, + "loss": 0.0334, + "step": 59600 + }, + { + "epoch": 1.8399283906413975, + "grad_norm": 0.34512731432914734, + "learning_rate": 1.1609853773164592e-08, + "loss": 0.0334, + "step": 59610 + }, + { + "epoch": 1.840237051669856, + "grad_norm": 0.3484075963497162, + "learning_rate": 1.1023614949751127e-08, + "loss": 0.0356, + "step": 59620 + }, + { + "epoch": 1.8405457126983147, + "grad_norm": 0.33167994022369385, + "learning_rate": 1.0452561457485744e-08, + "loss": 0.0342, + "step": 59630 + }, + { + "epoch": 1.8408543737267733, + "grad_norm": 0.2353219836950302, + "learning_rate": 9.896693469829688e-09, + "loss": 0.0327, + "step": 59640 + }, + { + "epoch": 1.841163034755232, + "grad_norm": 0.4828399121761322, + "learning_rate": 9.35601115564788e-09, + "loss": 0.0342, + "step": 59650 + }, + { + "epoch": 1.8414716957836903, + "grad_norm": 0.2777845859527588, + "learning_rate": 8.830514679186719e-09, + "loss": 0.0328, + "step": 59660 + }, + { + "epoch": 1.841780356812149, + "grad_norm": 0.3463110327720642, + "learning_rate": 8.320204200074066e-09, + "loss": 0.0323, + "step": 59670 + }, + { + "epoch": 1.8420890178406073, + "grad_norm": 0.25765857100486755, + "learning_rate": 7.825079873324814e-09, + "loss": 0.0321, + "step": 59680 + }, + { + "epoch": 1.842397678869066, + "grad_norm": 0.513580858707428, + "learning_rate": 7.345141849351977e-09, + "loss": 0.0334, + "step": 59690 + }, + { + "epoch": 1.8427063398975245, + "grad_norm": 0.3308010399341583, + "learning_rate": 6.880390273944493e-09, + "loss": 0.0325, + "step": 59700 + }, + { + "epoch": 1.8430150009259831, + "grad_norm": 0.31858518719673157, + "learning_rate": 6.4308252882838704e-09, + "loss": 0.0324, + "step": 59710 + }, + { + "epoch": 1.8433236619544418, + "grad_norm": 0.34215277433395386, + "learning_rate": 5.9964470289386455e-09, + "loss": 0.0325, + "step": 59720 + }, + { + "epoch": 1.8436323229829001, + "grad_norm": 0.41667258739471436, + "learning_rate": 5.577255627853273e-09, + "loss": 0.0319, + "step": 59730 + }, + { + "epoch": 1.8439409840113588, + "grad_norm": 0.37919551134109497, + "learning_rate": 5.173251212370334e-09, + "loss": 0.0309, + "step": 59740 + }, + { + "epoch": 1.8442496450398171, + "grad_norm": 0.2800273299217224, + "learning_rate": 4.784433905219432e-09, + "loss": 0.0328, + "step": 59750 + }, + { + "epoch": 1.8445583060682758, + "grad_norm": 0.2838169038295746, + "learning_rate": 4.4108038245060934e-09, + "loss": 0.0325, + "step": 59760 + }, + { + "epoch": 1.8448669670967344, + "grad_norm": 0.45092010498046875, + "learning_rate": 4.05236108373952e-09, + "loss": 0.0341, + "step": 59770 + }, + { + "epoch": 1.845175628125193, + "grad_norm": 0.35523295402526855, + "learning_rate": 3.7091057917937324e-09, + "loss": 0.0328, + "step": 59780 + }, + { + "epoch": 1.8454842891536516, + "grad_norm": 0.3302462697029114, + "learning_rate": 3.381038052946428e-09, + "loss": 0.0313, + "step": 59790 + }, + { + "epoch": 1.84579295018211, + "grad_norm": 0.5096606016159058, + "learning_rate": 3.0681579668623283e-09, + "loss": 0.0323, + "step": 59800 + }, + { + "epoch": 1.8461016112105686, + "grad_norm": 0.28964993357658386, + "learning_rate": 2.7704656285709727e-09, + "loss": 0.0354, + "step": 59810 + }, + { + "epoch": 1.846410272239027, + "grad_norm": 0.4036692976951599, + "learning_rate": 2.4879611285166803e-09, + "loss": 0.0329, + "step": 59820 + }, + { + "epoch": 1.8467189332674856, + "grad_norm": 0.3083502948284149, + "learning_rate": 2.2206445525085883e-09, + "loss": 0.0358, + "step": 59830 + }, + { + "epoch": 1.8470275942959442, + "grad_norm": 0.40798333287239075, + "learning_rate": 1.9685159817595112e-09, + "loss": 0.0345, + "step": 59840 + }, + { + "epoch": 1.8473362553244028, + "grad_norm": 0.22093266248703003, + "learning_rate": 1.7315754928470817e-09, + "loss": 0.0296, + "step": 59850 + }, + { + "epoch": 1.8476449163528614, + "grad_norm": 0.31972044706344604, + "learning_rate": 1.5098231577581611e-09, + "loss": 0.0336, + "step": 59860 + }, + { + "epoch": 1.8479535773813198, + "grad_norm": 0.38289061188697815, + "learning_rate": 1.3032590438499804e-09, + "loss": 0.036, + "step": 59870 + }, + { + "epoch": 1.8482622384097784, + "grad_norm": 0.45347267389297485, + "learning_rate": 1.1118832138723444e-09, + "loss": 0.0309, + "step": 59880 + }, + { + "epoch": 1.8485708994382368, + "grad_norm": 0.27819520235061646, + "learning_rate": 9.356957259620825e-10, + "loss": 0.0341, + "step": 59890 + }, + { + "epoch": 1.8488795604666954, + "grad_norm": 0.3221888244152069, + "learning_rate": 7.74696633637495e-10, + "loss": 0.0329, + "step": 59900 + }, + { + "epoch": 1.849188221495154, + "grad_norm": 0.4390029013156891, + "learning_rate": 6.288859858039064e-10, + "loss": 0.0344, + "step": 59910 + }, + { + "epoch": 1.8494968825236127, + "grad_norm": 0.3524657189846039, + "learning_rate": 4.982638267647665e-10, + "loss": 0.0362, + "step": 59920 + }, + { + "epoch": 1.8498055435520713, + "grad_norm": 0.27257341146469116, + "learning_rate": 3.8283019618834406e-10, + "loss": 0.0336, + "step": 59930 + }, + { + "epoch": 1.8501142045805297, + "grad_norm": 0.2837862968444824, + "learning_rate": 2.825851291410331e-10, + "loss": 0.0334, + "step": 59940 + }, + { + "epoch": 1.850422865608988, + "grad_norm": 0.2757164239883423, + "learning_rate": 1.975286560873535e-10, + "loss": 0.0352, + "step": 59950 + }, + { + "epoch": 1.8507315266374467, + "grad_norm": 0.2753726541996002, + "learning_rate": 1.2766080285109284e-10, + "loss": 0.032, + "step": 59960 + }, + { + "epoch": 1.8510401876659053, + "grad_norm": 0.31717827916145325, + "learning_rate": 7.298159065971533e-11, + "loss": 0.0346, + "step": 59970 + }, + { + "epoch": 1.851348848694364, + "grad_norm": 0.28027334809303284, + "learning_rate": 3.349103612770854e-11, + "loss": 0.0329, + "step": 59980 + }, + { + "epoch": 1.8516575097228225, + "grad_norm": 0.36368247866630554, + "learning_rate": 9.189151245481142e-12, + "loss": 0.0338, + "step": 59990 + }, + { + "epoch": 1.851966170751281, + "grad_norm": 0.3615170419216156, + "learning_rate": 7.594340156735769e-14, + "loss": 0.0335, + "step": 60000 + }, + { + "epoch": 1.851966170751281, + "step": 60000, + "total_flos": 0.0, + "train_loss": 0.04677600249151389, + "train_runtime": 91491.1365, + "train_samples_per_second": 41.971, + "train_steps_per_second": 0.656 + } + ], + "logging_steps": 10, + "max_steps": 60000, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 5000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 32, + "trial_name": null, + "trial_params": null +} diff --git a/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/training_args.bin b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..cef1c73165e1ec6f1cb6e65a6fd5465e0e6c1567 --- /dev/null +++ b/n1.5/vision-ft/Baseline_FullVision_Connector_FMOnly_VisionLR0p1_RoboCasaKitchen_N1d5_Full_BS64_60K/run_root/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dc9b6439676ac850023474c5b9cc2100b2a77fac22841b26a66e427c58a5e89c +size 5688